Professional Documents
Culture Documents
Operational Research Applied To Decisions in Home Health Care
Operational Research Applied To Decisions in Home Health Care
Operational Research Applied To Decisions in Home Health Care
To cite this article: Luca Grieco, Martin Utley & Sonya Crowe (2021) Operational research
applied to decisions in home health care: A systematic literature review, Journal of the
Operational Research Society, 72:9, 1960-1991, DOI: 10.1080/01605682.2020.1750311
ORIGINAL ARTICLE
CONTACT Luca Grieco l.grieco@ucl.ac.uk Department of Mathematics, Clinical Operational Research Unit, University College London, 4
Taviton Street, London WC1H 0BT, UK
Supplemental data for this article is available online at https://doi.org/10.1080/01605682.2020.1750311.
ß 2020 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits
unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY 1961
The authors point out that this undermines the operational decision problems addressed in the OR
potential benefits of improved operational decision literature, these workforce innovations include inter-
making by not making clear the strategic and tac- ventions to increase the use of information technol-
tical decisions required to create a context in which ogy to cut administration time and reduce the need
OR approaches to operational decisions can have for staff routes to start and/or end routes at a base,
most impact. introducing new scoring systems intended to quantify
The research questions raised by these previous workload more fairly, and considering the creation of
reviews include: new workforce roles. Some of these alter the nature
of operational problems addressed in the literature
What combinations and hierarchies of decisions and/or facilitate the implementation of OR solutions,
have been addressed in the OR literature on whereas others frame their own decision problems.
home health care and what aspects system per- This motivated us to bring to the fore the explicit
formance have been considered when addressing and implicit assumptions made in the literature about
each decision? workforce structure, terms of employment, preferen-
Can a group of research works be identified that, ces and agency. We consider how these underpinning
together, provide a comprehensive (in terms of assumptions may influence the feasibility, acceptabil-
the decisions addressed) and coherent (in terms ity or potential benefits associated with different OR
of the aspects of system performance considered) approaches in the context of the workforce arrange-
set of OR solutions to decision problems for ments that prevail in different health care systems.
Home Health Care? This work contributes to the body of operational
What decisions relevant to home health care and research on home health care by updating previous
amenable to OR have been given insufficient reviews, providing an analysis of the extent to which
attention in the literature? the current OR literature provides a basis for
improved, coherent decision making across the
In this paper we address these research ques- spectrum of decisions faced in home health care,
tions by: and identifying priority areas for future research, in
terms of decision problems to address and context-
counting the numbers of papers that address dif- ual and environmental features of home health care
ferent combinations and hierarchies of decision;
systems that should be reported alongside model
reviewing how the performance of HHC systems
formulations and solution approaches.
is conceived within the literature and mapping
the different aspects of system performance con-
sidered within the literature to the different deci- 2. Methods
sions addressed; 2.1. Decision problems in home health care
clustering papers by the aspects of performance
considered and identifying the sets of decisions We identified an initial set of decisions at different
spanned within each cluster; planning levels relevant to the design and delivery of
comparing the set of decisions addressed in the HHC from Hulshof et al. (2012). We split some of
literature to a hierarchical framework of decisions these into multiple decisions, for example Hulshoff’s
in HHC compiled with input from discussions ‘Visit (re-)scheduling’ was separated into four discrete
with HHC decision makers in North decisions, ‘allocation of staff’, ‘visit scheduling’, ‘routing
East London. of visits’ and ‘short-term fixes’, whilst acknowledging
that sometimes these decisions are combined within
As a secondary research question, we sought to modelling approaches in the literature.
identify OR approaches used outside HHC that could We then broadened this based on discussions
be adapted to address decision problems neglected by with HHC stakeholders in North East London,
the set of papers identified in this review. which comprised 29 meetings and 3 workshops with
Our review has been undertaken as part of a study people working in the HHC system, including man-
in which we are working with a number of home agers and frontline nursing staff from the local
health care organisations. While these organisations home health provider and commissioners of home
face problems of districting, service definition, con- health services, managers of local voluntary organi-
tract design and capacity planning, their efforts to sations providing support to home care patients and
improve efficiency focus largely on what they term as their families and a group of service user representa-
“workforce innovations”, essentially changes to the tives. In relation to the set of decisions considered
work done by clinicians and how this work is organ- in this paper, our findings from these discussions
ised. As well as including interventions to tackle led us to augment the frameworks of home health
1962 L. GRIECO ET AL.
care decisions presented elsewhere in the literature In addition to incorporating these findings into
in the following ways: our framework, we identified logical links and hier-
archical orderings between the decisions to ensure
Stakeholders are currently considering the intro- consistency with existing descriptions of HHC from
duction of new workforce roles that blend skills a supply chain perspective (Gutierrez & Vidal, 2013;
associated with multiple roles, including across Sahin & Matta, 2015). We then devised a graphical
different traditional professional disciplines (e.g. representation of this decision framework (Figure
hybrid nursing and physio support roles). We 1), in which we placed the output of a home visit
therefore felt that defining workforce roles was a on the right hand side of the page and then worked
decision that needed to be differentiated from back through the decisions (and other inputs) neces-
that of resource dimensioning or allocation (as sary for a specific professional to visit a specific
used by Hulshof et al., 2012), which generally patient at a particular time on a particular day. This
focus on determining the number of staff needed led to a group of operational decisions being placed
from existing roles. in the centre of the page. Within this left-to-right
The development of home health care plans for flow of quasi-precedence relationships, we chose to
patients is not explicitly covered by Hulshof display organisations’ prior strategic and tactical
et al. (2012) or previous reviews on HHC (e.g. decisions as feeding into operational decisions from
Gutierrez & Vidal, 2013; Sahin & Matta, 2015 above, with prior patient-specific decisions feeding
and Cisse et al., 2017), but we found it to be an into operational decisions from below. Both of these
important decision that stakeholders face, which strands were considered to be dependent on a deci-
relates not only to operational decisions about sion around the definition of service to be offered
scheduling visits but also to tactical decisions and the terms of access for patients, with this placed
regarding team size and composition. Related to at the far left of the page.
this, the timing and nature of reassessments of Where possible, our working definition of each
patient need, which feed into amendments to decision was made consistent with the nomenclature
care plans, was also considered an important used in Hulshof et al. (2012).
decision by stakeholders (and absent from
Hulshof et al., 2012) and the frameworks pre-
sented in the previous reviews). 2.2. Literature search
Many patients receive multiple home-based 2.2.1. Operational research applied to home
health or care services and informal support, and health care
stakeholders felt it was important to consider We conducted a systematic review to assess the
how these services can coordinate effectively. extent to which the development and adoption of
Although Sahin and Matta (2015) refer to deci- OR approaches address the identified set of HHC
sions about partnership strategy at a strategic decision problems.
level and the identification of partners at a tac- We searched the electronic database Web of
tical level, they do not make explicit coordination Science (https://wok.mimas.ac.uk/) using a combin-
with other services at an operational level, and ation of the search terms listed in Appendix 1
this is also absent in Hulshoff’s taxonomy (Tables A1 and A2), and extracting records with:
(Hulshof et al., 2012) and the other HHC review
articles (Gutierrez & Vidal, 2013; Cisse at least one of the terms related to OR methods
et al., 2017). set out in Table A1 in the article title, abstract or
There are important patient-level service proc- keywords (informed by Palmer et al. 2018);
esses that feed into the decisions, such as identi- AND
fying and specifying a patient’s need(s), and at least one of the terms related to the setting of
some stakeholders focus primarily on the main HHC set out in Table A1 in the article title,
patient-level output from the decisions, namely abstract or keywords;
the delivery of a home visit. Therefore, in line AND
with Gutierrez and Vidal (2013), who include an at least one of the decision terms set out in Table
explicit dimension on service processes in their A2 in the article title, abstract or keywords.
framework, we added the decision inputs ‘need
identified’ and ‘referral’ to our framework along Note that our choice of search terms respected
with the decision output ‘home visit’. We also that the distinction between health care and social/
made explicit the distinction between decisions personal care activities can differ among systems,
and their inputs and outputs. meaning that some of the papers identified relate to
JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY 1963
Figure 1. A graphical representation of the set of decision problems identified as associated with the design and operation of
a home health care service. The framework diagram shows a flow of decisions from the definition of the service to delivery of
care at a patient’s home, with some key inputs related to the patient shown at the bottom.
the delivery of personal social care by non-health 2.2.2. Operational research applied to analogous
professionals. decision problems
We considered all publications in English with We conducted an additional scoping review of OR
no constraints on publication type or year. The approaches to decisions analogous to those in HHC
search was run on the 28th September 2018. All that have been developed for other settings outside
publications found were then assessed against this HHC. To do this, we used the same search strategy
sequence of criteria: described in 2.2.1 but without the “Settings” terms
and with some HHC specific “Decision” terms
Peer-reviewed: The titles and abstracts of only adapted (for instance “nurse” was replaced with
peer-reviewed articles were assessed. “worker” in searches – see Table A3 in Appendix 1
Title/abstract: Review articles were subject to full for full details). Defined this way, analogous deci-
text reading if the title and/or the abstract sug- sions included, for example, the scheduling and
gested that a subset of the papers reviewed focused routing of workers providing maintenance services
on tackling decisions in HHC through OR at a number of client locations or the size and com-
approaches. Other publication types (e.g. research position of teams in other industries. Because of this
articles, peer-reviewed conference proceedings, broad definition and the anticipated volume of
book chapters) were subject to full text reading if papers related to some decisions, we restricted this
the title and/or abstract suggested that the work scoping review to review articles. From the search
focused on tackling one or more of the decisions results associated with each analogous decision, we
of interest in HHC through an OR approach. studied the two most recent review articles describ-
Full text: Articles were retained if the above criter- ing decision problems where we considered the
ion was still satisfied on reading of the full text structure of the problem and the nature of objec-
and, for non-review articles, if the description of tives/requirements to be similar to those we found
the approach was sufficiently detailed to identify in HHC settings.
modelling assumptions, formulation and outputs.
2.3. Data extraction
Additional research papers identified from
accepted review articles were then subject to the The study team agreed on an initial data extraction
same sequence of inclusion criteria. We only table for the systematic review. Data extracted
extracted data from non-review article publica- included the aim of the study, the HHC decisions
tion types. addressed, the planning horizon considered, the
1964 L. GRIECO ET AL.
modelling and solution approaches taken, the decisions. To do this, we adapted a clustering algo-
aspects of performance considered (and associated rithm to find groups of papers that were coherent
metrics) and the level of engagement with practice. in terms of how system performance was conceived
One reviewer (LG) carried out the article search and and then looked at how comprehensive the set of
selection, reading and data extraction, making an ini- decisions addressed in each cluster was. To explore
tial assessment as to the decision(s) in HHC the coherence of papers in terms of the aspects of
addressed. In cases where the same aspect of perform- system performance considered, we computed a
ance was labelled differently in different publications similarity score si, j 2 ½0, 1 between all possible pairs
(e.g. “caregiver utilisation”, “worker utilisation”, of papers i and j:
“resource utilisation”), we chose a label (e.g. “staff uti- A i \ Aj
lisation”) for our analyses. For papers in which the si, j :¼
Ai [ Aj
aspects of HHC system performance being promoted
were not explicit, a judgement was made based on the with Ak denoting the set of aspects of system per-
objective functions and constraints of models and on formance considered in paper k (the higher the
the presentation of results. overlap between papers i and j, the closer si, j to 1).
To identify and to try to minimise the impact of We then built an undirected, weighted graph with
subjective assessments in extracting data, a second each node denoting a paper and an arc with weight
reviewer (MU) read 20% of the selected papers si, j existing between node i and node j if si, j 0:6:
sampled at random but with at least one paper per We identified each connected component (i.e. set of
decision where possible. Data extracted by the nodes in the graph that are linked to each other by
second reader was compared with that obtained paths) in the obtained graph as a cluster of papers
from first reading. Any systematic difference in dealing with similar sets of aspects of system per-
extraction or interpretation of data was resolved formance. Note that the lower the cut-off, the higher
through discussion with the senior author (SC). the average size of the connected components in the
This process led to the iterative refinements of the graph. We chose the cut-off (0.6) as the minimum
data extraction table allowing us to capture import- real number such that at least two of the obtained
ant distinctions between papers relevant to our clusters were formed of more than four papers.
research questions. For instance, we refined our
mapping of papers to HHC decisions to specify 3. Results
whether papers focused on and proposed a quantita-
tive model for a given decision, considered that 3.1. Decision framework
decision as a secondary outcome, made explicit Figure 1 shows the set of decisions identified as
assumptions about that decision as a model input, important to the design and delivery of HHC serv-
or simply made reference to that decision. Finally, ices, spanning the strategic, tactical and operational
the first reviewer revisited the completed data planning levels. There is an implied hierarchy, or
extraction table to ensure a consistent presentation logical precedence relationship between some deci-
and level of detail across all entries. sions, whereas others may be made in combination
or in different sequences.
2.4. Analysis
For each paper, we identified the partial hierarchy 3.2. Literature search output
of decisions in HHC considered. We then counted 3.2.1. Summary of articles retrieved and data
the number of papers that considered each unique extracted
hierarchy and produced a graphical summary of Our search identified 222 articles, with 4 review
these data based on the graphical representation of articles and 62 non-review articles meeting our
the decision framework devised at section 2.1. For inclusion criteria (Figure 2). Fifteen additional non-
each decision, we then tabulated the OR approaches review articles identified in the 4 review articles
and accompanying solution methods used in the lit- were accepted. Table 1 shows the data extracted
erature and, separately, the aspects of system per- from the 77 non-review articles.
formance considered.
3.2.2. Decision hierarchies
2.4.1. Paper clustering Among the 59 of the 77 papers that addressed more
One of our aims was to identify whether there are than one decision, we identified 15 distinct partial
groups of papers that, between them, provide a hierarchies of decision (see Figure 3). The most
comprehensive and coherent basis for decision mak- commonly considered partial hierarchy (30 papers)
ing across strategic, tactical and operational was where staff rosters and home health care plans
JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY 1965
Figure 2. Literature search flow diagram, showing the number (N) of articles identified, meeting the inclusion criteria of peer-
review, and retained for analysis after title/abstract assessment and after full-text reading. The left-hand side of the diagram
shows the flow for non-review articles, whilst the right-hand side shows our process for identifying additional research papers
from accepted review articles (which were then subject to the same sequence of inclusion criteria). We only extracted data
from non-review article publication types.
Table 1. Data extracted from the non-review article publications that met the selection criteria.
1966
An et al. (2012) Mixed-integer programming Visit scheduling Multi-period Mixed-integer Two-phase heuristic algorithm: i) Scheduler productivity (Planning Synthetic instances
formulation and solution Routing of visits (days) programming determine a partial schedule for time) with realistic
approach for the problem of patients with shorter care Staff travelling (Travelling time) features
scheduling visits for a nurse intervals; ii) add remaining
over a time horizon to fullfil patients to obtain a
the needs of a set of already complete schedule
assigned patients.
Argiento Bayesian framework to NA Multi-period General linear mixed Markov Chain Monte Continuity of care (No stated Synthetic instances
et al. (2016) represent patients’ home (weeks) model with Carlo simulation metrics of performance) with realistic
care demand evolution with parameter Staff utilisation (No stated features
time and to predict demand estimation through metrics of performance)
in future periods. Markov Chain
Monte
Carlo simulation
Bachouch Integer programming Staff rostering Multi-period Integer programming Not specified - commercially Continuity of care (Staff per None
et al. (2011) formulation for the rostering, Visit scheduling (days) available solver (LINGO and patient)
scheduling, allocation and Allocation of staff CPLEX) used Scheduler productivity (Planning
routing problems in Routing of visits time)
health care. Staff travelling (Travelling time)
Bard Development and comparison Allocation of staff Multi-period Mixed-integer The first two models were easily Costs of staff time spent with Instances obtained
et al. (2013) of a set of integer Routing of visits (days) programming solved using commercial solvers patient (Treatment costs) from organisations
programming models and (CPLEX). For the third model, Number of teams (Number of
heuristic methods to allocate two heuristic algorithms were home bases)
and route therapists to visits developed that determine Staff travelling (Travelling costs)
on a weekly basis. feasible solutions starting from Team size (Number of staff in
lower bounds computed by a team)
solving relaxed versions of the
original problem.
Bard, Shao and Heuristic algorithm for the Staff rostering Multi-period Mixed-integer Heuristic approach based on a Costs of staff time spent with Instances obtained
Jarrah (2014) problem of constructing Visit scheduling (days) programming greedy randomised adaptive patient (Treatment costs) from organisations
weekly tours for home Allocation of staff search procedure Scheduler productivity (Planning
care therapists. Routing of visits time)
Staff productivity (Ratio between
treatment cost and total cost)
Staff travelling (Travelling costs)
Bard, Shao, Qi, Mixed-integer programming Visit scheduling Multi-period Mixed-integer Branch-cut-and-price approach Costs of staff time spent with Instances obtained
et al. (2014) formulation and two Allocation of staff (days) programming Heuristic algorithm based on patient (Treatment costs) from organisations
solution algorithms for the Routing of visits rolling horizon method Scheduler productivity (Planning
problem of constructing time)
weekly tours for home Staff productivity (No stated
care therapists. metrics of performance)
Staff travelling (Travelling costs)
(continued)
Table 1. Continued.
Planning Aspects of performance (metrics of Level of engagement
ID Aim of the study Decisions modelled horizon Modelling approach Solution approach performance) with current practice
Begur et al. Development of a decision Allocation of staff Single-period Mixed-integer Clarke and Wright’s savings Scheduler productivity (Planning Actual collaboration
(1997) support system for Routing of visits (day) programming algorithm for staff-visit time) with organisations
scheduling and routing Short-term fixes allocation þ Rosenkrants, Sterns Staff productivity with
nurses at a home health and Lewis’s algorithm for route (Travelling time) implementation
care provider embedding a refinement of local travelling efforts
database management salesman sub-problems
system, a heuristic algorithm
and GIS technology.
Bennett and Two heuristic approaches to Visit scheduling Multi-period Combinatorial Greedy heuristic approaches: Scheduler productivity (Planning None
Erera (2011) solve the dynamic periodic Routing of visits (weeks) optimisation distance-based insertion time)
fixed appointment time criterion Staff productivity (Number of
scheduling and capacity-based visits)
routing problem. insertion criterion Staff travelling (Travelling time
per visit)
Benzarti Two mixed-integer Districting Single-period Mixed-integer Not specified - commercially Continuity of care (No stated None
et al. (2013) programming formulations (several years) programming available solver (CPLEX) used metrics of performance)
for the districting problem in District compactness (Intra-
home health care district distances)
Staff travelling (Intra-district
distances)
Team size (No stated metrics of
performance)
Workload balance (Deviation
from average workload)
Bertels and Mixed-integer programming Visit scheduling Single-period Mathematical Combination of constraint Match of staff skills to patient None
Fahle (2006) formulation and solution Allocation of staff (day) programming programming, tabu search and needs (Mismatch between
method combining the Routing of visits formulation solved simulated annealing allocated staff skills and
rostering and routing using a required skills)
problems in home combination of Patient preferences (Deviation
health care. constraint from preferred visit time)
programming, tabu Scheduler productivity (Planning
search and time)
simulated Staff preferences (Deviation from
annealing. planned shift)
Staff travelling (Travelling time)
Blais Non-linear programming model Districting Single-period Non-linear Tabu search heuristic developed Scheduler productivity (Planning Actual collaboration
et al. (2003) to re-partition a territory (several programming by Bozkaya et al. for political time) with organisations
from 4 to 6 districts in order years) (apparently based on districting Staff travelling (Travelling time) with
to alleviate difficulties in a set covering Workload balance (Deviation implementation
managing nursing teams. formulation, though from average workload) efforts
constraints are not
explicitly reported)
Bowers Solution algorithm for the Visit scheduling Multi-period Combinatorial Modified Clarke-Wright algorithm Continuity of care (Deviation None
et al. (2015) problems of scheduling, Allocation of staff (days) optimisation including a component in the from preferred staff, Number of
JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY
allocation and routing over a Routing of visits "savings" to reflect the benefit handovers)
time horizon, applied of the chosen staff matching Patient preferences (Deviation
to midwifery. patient’s preferences from preferred staff)
Staff travelling (Travelling time)
1967
(continued)
Table 1. Continued.
1968
Cappanera and Integer linear programming Visit scheduling Multi-period Integer linear Branch and bound (using CPLEX) Scheduler productivity (Planning Synthetic instances
Scutella formulations and heuristic Allocation of staff (days) programming time) with
(2015) procedure to jointly address Routing of visits Staff travelling (Maximum staff realistic features
the scheduling, allocation utilisation rate)
and routing problems in Workload balance (Minimum
home care. staff utilisation rate)
Cappanera Cardinality-constrained approach Visit scheduling Multi-period Cardinality-constraint Metaheuristic approach Continuity of care (Staff per Synthetic instances
et al. (2018) to address the scheduling, Allocation of staff (days) robust approach patient) with
allocation and routing Routing of visits Scheduler productivity (Planning realistic features
problems in home health care, time)
with patient demand subject Workload balance (Maximum
to uncertainty. staff utilisation rate)
Carello and Cardinality-constrained approach Allocation of staff Multi-period Cardinality-constrained Not specified - commercially Continuity of care (Reallocation Instances obtained
Lanzarone to address the problem of (weeks) approach (robust available solver (CPLEX) used costs) Overtime from organisations
(2014) allocation of staff to patients in optimisation) (Overtime costs)
home health care, with patient
demand subject to uncertainty.
Carello Cardinality-constrained Allocation of staff Multi-period Integer linear Commercial solver (CPLEX) Continuity of care (Staff switches Synthetic instances
et al. (2018) approach to address the (weeks) programming per patient) with
problem of allocating staff Overtime (Overtime costs) realistic features
to patients in home health Scheduler productivity (Planning
care, with patient demand time)
subject to uncertainty and Workload balance (Maximum
accounting for perspectives staff utilisation rate)
of different stakeholders.
Cattafi Constraint programming Visit scheduling Multi-period Constraint Heuristic approach based on large Continuity of care (Staff per patient) Instances obtained
et al. (2015) formulation and heuristic Allocation of staff (days) programming neighbourhood search Scheduler productivity (Planning from organisations
approach for the problems Routing of visits time)
of scheduling and routing in Staff travelling (Travelling time)
home health care. Workload balance (Staff workload)
De Stochastic programming model Commissioning Multi-period Markov chain and Not mentioned - however, it Budget sufficiency (Number of Synthetic instances
Angelis to determine the best Screening (weeks) stochastic linear seems that the linear accepted patients) with
(1998) resource allocation to home programming approximation used makes the Staff utilisation (Number of realistic features
care for AIDS patients, stochastic problem easy accepted patients)
within a limited budget and to solve
providing a minimum
standard of service.
Decerle et al. Mixed-integer programming Visit scheduling Single-period Mixed-integer Two-phase meta- Costs of staff time spent with None
(2016) formulation and heuristic Allocation of staff (day) programming heuristic approach patient (Treatment costs)
algorithm for the problems Routing of visits Scheduler productivity (Planning
of scheduling, allocation and Coordination with other time)
routing in home health care. home care services Staff travelling (Travelling costs)
(continued)
Table 1. Continued.
Planning Aspects of performance (metrics of Level of engagement
ID Aim of the study Decisions modelled horizon Modelling approach Solution approach performance) with current practice
Decerle Mixed-integer programming Visit scheduling Single-period Mixed-integer Memetic algorithm Patient preferences (Deviation Synthetic instances
et al. (2018) formulation and heuristic Allocation of staff (day) programming from preferred visit time) with
algorithm for the problems Routing of visits Staff travelling (Travelling realistic features
of scheduling, allocation and Coordination with other costs)
routing in home health care. home care services Synchronisation of visits (Penalty
for non-respected
synchronisation of visits)
Du et al. (2017) Integer programming Visit scheduling Single-period Integer linear Genetic algorithm with Costs of staff time spent with None
formulation and heuristic Allocation of staff (day) programming local search patient (Treatment costs)
algorithm for the problems Routing of visits Patient preferences (Deviation
of scheduling, allocation and from preferred visit time)
routing in home health care. Staff travelling (Travelling costs)
En-Nahli Multi-objective mixed-integer Visit scheduling Single-period Multi-objective Not specified - commercially Patient preferences (Deviation None
et al. (2015) programming formulation Allocation of staff (day) optimisation available solver (CPLEX) used from preferred staff)
for the problems of Routing of visits (mixed-integer Staff preferences (Minimum staff
scheduling, allocation and programming) utilisation rate, Staff waiting
routing in home health care. time)
Staff travelling (Travelling time)
Erdem and Two-stage heuristic solution Visit scheduling Single-period Mixed-integer Two-stage algorithm: i) clustering Demand satisfaction (Penalty for None
Bulkan approach for large-scale Allocation of staff (day) programming of nurses and jobs; ii) unassigned visits)
(2017) instances of the problem of Routing of visits generation of scheduling/ Overtime (Overtime costs)
scheduling, allocation and routing using a variable Staff preferences (Deviation from
routing in home health care. neighbourhood preferred visit time)
search algorithm Staff travelling (Travelling costs)
Errarhout Two-stage stochastic Allocation of staff Multi-period Stochastic, mixed- Not specified - commercially Staff travelling (Travelling None
et al. (2016) programming approach for (weeks) integer available solver (CPLEX) used distance)
the problem of allocating programming Workload balance (Workload
nurses to patients over a variability)
time horizon in home
health care.
Eveborn Decision support system aimed Visit scheduling Multi-period Integer programming Repeated matching (combination Continuity of care (Deviation Actual collaboration
et al. (2006) at reducing the time spent Allocation of staff (days) (based on set of optimisation methods from staff working area, Staff with organisations
planning daily visits by Routing of visits partitioning model) and heuristics) per patient) with
home health care Short-term fixes Match of staff skills to patient implementation
organisations as well as at needs (Mismatch between efforts
improving service efficiency. allocated staff skills and
required skills)
Patient preferences (Deviation
from preferred staff,
Deviation from preferred
visit time)
Scheduler productivity (Planning
time)
Staff preferences (Deviation from
JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY
planned shift)
Staff travelling (Travelling costs,
Travelling time)
(continued)
1969
Table 1. Continued.
1970
required skills)
Patient preferences (Deviation
from preferred staff, Deviation
from preferred visit time)
Scheduler productivity (Planning
time)
Staff preferences (Deviation from
planned shift)
Staff travelling (Travelling costs,
Travelling time)
Fikar and Integer programming Visit scheduling Single-period Integer programming Matheuristic approach based on Overtime (Deviation from Synthetic instances
Hirsch (2015) formulation and heuristic Allocation of staff (day) tabu search and consisting of planned shift) with
algorithm for the problems Routing of visits two stages: creation of walking realistic features
of scheduling, allocation and routes and optimisation of the
routing in home health care. transport system
Fikar Discrete-event driven heuristic Visit scheduling Single-period Integer programming Discrete-event driven Scheduler productivity (Planning None
et al. (2016) approach for the problems Allocation of staff (day) metaheuristic approach time)
of scheduling, allocation and Routing of visits Staff utilisation (Staff utilisation
routing in home health care. Short-term fixes rate, Vehicle utilisation rate)
Guericke and Mixed-integer programming Staff rostering Multi-period Mixed-integer Meta-heuristic approach based on Staff utilisation (Penalty for None
Suhl (2017) formulation and heuristic Visit scheduling (days) programming adaptive large unassigned visits, Penalty for
algorithm for the problems Allocation of staff neighbourhood search unassigned weekend shifts,
of rostering, scheduling, Routing of visits Staff waiting time,
allocation and routing in Travelling time)
home health care.
Hertz and Mixed-integer programming Allocation of staff Single-period Mixed-integer Meta-heuristic approach based on Workload balance (Deviation None
Lahrichi formulation and heuristic (month) programming Tabu search from average number of
(2009) algorithm for the problem of patients per staff)
allocating nurses to patients
in home health care, with
dynamic assignment based
on current workload.
Hewitt Using an existing integer Staff rostering Multi-period Mixed-integer Heuristic approach based on the Continuity of care (Staff per None
et al. (2016) programming formulation Visit scheduling (days) programming "record-to-record patient)
for the problems of Allocation of staff travel algorithm" Staff travelling (Travelling time)
rostering, scheduling, Routing of visits
allocation and routing in
home health care in order to
compare different
planning horizons.
(continued)
Table 1. Continued.
Planning Aspects of performance (metrics of Level of engagement
ID Aim of the study Decisions modelled horizon Modelling approach Solution approach performance) with current practice
Hiermann Generic framework for the Visit scheduling Single-period Two-stage approach Initial solution determined Demand satisfaction (Penalty for Synthetic instances
et al. (2015) problems of scheduling, Allocation of staff (day) based on constraint through constraint unassigned visits) with
allocation and routing in Routing of visits programming and programming or random Overtime (Deviation from realistic features
home health care that can metaheuristics construction planned shift)
be easily adapted to include Solution improvement through a Patient preferences (Deviation
different requirements. metaheuristic algorithm: from preferred staff, Deviation
variable neighbourhood search, from preferred visit time)
memetic algorithm, scatter Scheduler productivity (Planning
search, simulated annealing time)
Staff preferences (Deviation from
planned shift, Deviation from
preferred patient, Travelling
time)
Staff travelling (Travelling time)
Hindle Estimating yearly travelling NA Single-period Multiplicative NA Demand satisfaction (Travelling Actual collaboration
et al. (2000) distance for home health (year) analytical model distance, Travelling time) with organisations
care to inform with
fund allocation. implementation
efforts
Hindle Estimating yearly travelling NA Single-period Simplified Modelling NA Demand satisfaction (Travelling Instances obtained
et al. (2009) distance for home health (year) of Spatial Systems, distance, Travelling time) from organisations
care to inform consisting of using
fund allocation. simple
mathematical
functions to
approximate
travelling distances
and times to
degrees of accuracy
comparable to
those achieved by
more complex
routing algorithms
Koeleman Markov decision process to Screening Single-period Markov Analytical solution of the model Demand satisfaction (Cost of None
et al. (2012) derive optimal patient Team size and (several decision process Heuristic approach ("trunk accepting patients, Cost of
admission policies and composition weeks) reservation") delaying patients, Cost of
workforce size in home care. rejecting patients)
Lanzarone Markov chains to support NA Multi-period Stochastic modelling NA Continuity of care (No stated Actual collaboration
et al. (2010) resource planning in home (weeks) (Markov chain) þ metrics of performance) with organisations
care by estimating workload statistical data Workload balance (No stated with
over a time horizon. analysis metrics of performance) implementation
(not specified) efforts
Lanzarone and Stochastic programming Allocation of staff Single-period Stochastic modelling Analytical derivation of the best Continuity of care (Staff per Synthetic instances
Matta (2012) approach for allocating (several assignment policy patient) with
JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY
Lin et al. (2016) Mixed-integer programming Allocation of staff Multi-period Mixed-integer Not specified - commercially Demand satisfaction (Number of Synthetic instances
formulation for the problem (days) programming available solver (Gurobi allocated patients) with realistic
of assigning therapists to Optimizer) used features
patients over a time horizon
in home health care.
Lin et al. (2017) Mixed-integer programming Districting Single-period Mixed-integer Heuristic method based on a Number of teams (Number of None
formulation and heuristic (day) programming greedy approach districts)
algorithm for the districting Scheduler productivity
problem in the context of the (Planning time)
"meals-on-wheels" service.
Lin et al. (2018) Mixed-integer programming Staff rostering Multi-period Mixed-integer Meta-heuristic approach based on Disruptions to planned work None
formulation and heuristic Visit scheduling (days) programming harmony search (Deviation from original visit
algorithm for the problems Allocation of staff schedule)
of rostering, scheduling, Routing of visits Overtime (Overtime costs)
allocation, routing and re- Short-term fixes Staff travelling (Travelling costs)
rostering in home
health care.
Liu et al. (2013) Mixed-integer programming Visit scheduling Single-period Mixed-integer Tabu search heuristic Staff travelling (Travelling costs) None
formulation, tabu search Allocation of staff (day) programming Genetic algorithm
algorithm and genetic Routing of visits
algorithm for the problems
of scheduling, allocation and
routing in home health care.
Liu et al. (2014) Tabu search algorithm for the Visit scheduling Multi-period Mixed-integer Tabu search heuristic Staff travelling None
problems of scheduling, Allocation of staff (days) programming (Travelling distance)
allocation and routing in Routing of visits
home health care.
Liu et al. (2017) Branch and price algorithm for Visit scheduling Single-period Integer linear Branch-and-price algorithm Demand satisfaction (Penalty for None
the problems of scheduling, Allocation of staff (day) programming unassigned visits)
allocation and routing in Routing of visits Scheduler productivity (Planning
home health care. time)
Staff travelling (Travelling costs)
Liu et al. (2018) Bi-objective mixed-integer Visit scheduling Multi-period Bi-objective mixed- Metaheuristic approach Costs of staff time spent with None
programming approach for Allocation of staff (days) integer patient (Treatment costs)
the problems of scheduling, Routing of visits programming Patient preferences (Additional
allocation and routing in treatment time)
home health care. Staff travelling (Travelling costs)
opez-Santana
L Multi-agent approach for the Visit scheduling Multi-period Multi-agent approach, JADE (Java Agent Development Demand satisfaction (Amount of None
et al. (2016) problems of scheduling, Allocation of staff (day fractions) a step of which Framework), with MIP problem skills allocated to patients)
allocation and routing in Routing of visits consists of a solved using commercial Staff travelling (Travelling time)
home health care, enabling Short-term fixes mixed-integer optimisation software
dynamic handling of programming
new requests. problem
(continued)
Table 1. Continued.
Planning Aspects of performance (metrics of Level of engagement
ID Aim of the study Decisions modelled horizon Modelling approach Solution approach performance) with current practice
Manerba and Branch-and-cut approach to Visit scheduling Multi-period Integer programming Branch-and-price algorithm Match of staff skills to patient None
Mansini solve the problems of Allocation of staff (days) needs (Suitability score of
(2016) scheduling, allocation and Routing of visits patient-staff allocation)
routing in home health care.
Mankowska Mixed-integer programming Visit scheduling Single-period Mixed-integer The authors present several Patient preferences (Deviation None
et al. (2014) formulation and heuristic Allocation of staff (day) programming heuristic approaches that can from preferred visit time)
algorithms for the problems Routing of visits deal with the complexity of Punctuality (Tardiness)
of scheduling, allocation and Coordination with other the problem Staff travelling
routing in home health care. home care services (Travelling distance)
Marcon Multi-agent simulation model for Short-term fixes Single-period Multi-agent modelling Multi-agent simulation Staff autonomy (No stated None
et al. (2017) dynamically solving the routing (day) metrics of performance)
problem in home health care, Staff travelling (Travelling time)
given a pre-existing scheduling,
allocation and routing solution.
Maya Duque Bi-objective integer Visit scheduling Multi-period Bi-objective Three-stage process of identifying Patient preferences (Deviation Actual collaboration
et al. (2015) programming formulation Allocation of staff (days) optimisation set of possible schedules for a from preferred visit time, with organisations
and heuristic approach for Routing of visits (integer patient, set partitioning to Suitability score of patient-staff with
the problems of scheduling, programming), allocate staff to patient allocation) implementation
allocation and routing, with formulation schedules, then randomised Scheduler productivity (Planning efforts
presented as the core based on set local search to reduce travel time)
optimisation component of a partitioning time within permit margin on Staff preferences (Deviation from
decision support system preference meeting preferred visit time, Suitability
developed to the benefit of score of patient-staff allocation)
a home care organisation. Staff travelling
(Travelling distance)
Mutingi and Genetic algorithm applied to Visit scheduling Single-period Combinatorial Group genetic algorithm Patient preferences (Deviation None
Mbohwa routing of visits in home Allocation of staff (day) optimisation from preferred visit time)
(2013) health care. Routing of visits Scheduler productivity (Planning
time)
Staff travelling (Travelling
distance)
Staff utilisation (No stated
metrics of performance)
Mutingi and Fuzzy-based particle swarm Visit scheduling Single-period Integer programming Fuzzy-based particle swarm Patient preferences (Deviation None
Mbohwa optimisation approach for Allocation of staff (day) optimisation algorithm from preferred visit time)
(2014) the problem of assigning Routing of visits Scheduler productivity (Planning
staff to visits in home time)
health care. Workload balance (Deviation
from average visits per staff)
Nasir and Mixed-integer programming Visit scheduling Single-period Mixed-integer Not specified - commercially Patient preferences (Deviation None
Dang (2016) formulation for the problems of Allocation of staff (day) programming available solver (CPLEX) used from preferred visit time)
scheduling, allocation and Routing of visits Staff travelling (Travelling costs)
routing in home health care
and method to predict whether
JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY
(continued)
Table 1. Continued.
1974
allocation and routing in Routing of visits from average staff idle time,
home health care. Workload variability)
Nasir Integer programming Team size and Single-period Integer linear Commercial solver (CPLEX) Match of staff skills to patient None
et al. (2018) formulation for the problem composition (day) programming needs (Penalty for allocating
of determining the optimal Visit scheduling staff without experience of that
number of staff to deploy Allocation of staff patient type)
and offices to open over a Coordination with other Patient preferences (Score of
given time window, with the home care services similarity of patients belonging
aim to enable coordination to the same cluster)
between home health care Staff costs (Wage costs)
and telehealth services. Staff travelling (Travelling costs)
Nickel Constraint programming Staff rostering Multi-period Constraint Constraint programming Continuity of care (Staff per Synthetic instances
et al. (2012) formulations and heuristic Visit scheduling (days) programming techniquest þ heuristics on patient) with
approaches for the problems Allocation of staff adaptive large neighbourhood Demand satisfaction (Penalty for realistic features
of rostering, scheduling, Routing of visits search and tabu search unassigned visits)
allocation and routing in Short-term fixes Disruptions to planned work
home health care. (Deviation from original visit
schedule)
Overtime (Deviation from
planned shift)
Staff travelling
(Travelling distance)
Quintana Clustering technique for solving Visit scheduling Multi-period Clustering Heuristic method based on a Staff productivity (Number of Instances obtained
et al. (2017) very large instances of the Allocation of staff (days) greedy approach staff deployed) from organisations
problems of scheduling, Routing of visits Staff utilisation
allocation and routing in (Unproductive time)
home care, with the aim to
find the best trade-off
between the number of
caregivers deployed and the
amount of
unproductive time.
Rasmussen Set partitioning formulation Visit scheduling Single-period Integer programming Branch-and-price algorithm, Demand satisfaction (Penalty for Synthetic instances
et al. (2012) and branch and price Allocation of staff (day) coupled with a procedure of unassigned visits) with
algorithm for the problems Routing of visits visit clustering Patient preferences (Deviation realistic features
of scheduling, allocation and from preferred patient)
routing in home care. Staff travelling (Travelling costs)
Redjem and Heuristic algorithm for the Visit scheduling Single-period Mixed-integer Heuristic approach named Scheduler productivity (Planning Instances obtained
Marcon problems of scheduling and Routing of visits (day) programming "Caregivers Routing Heuristic" time) from organisations
(2016) routing in home care. Staff productivity
(Unproductive time)
(continued)
Table 1. Continued.
Planning Aspects of performance (metrics of Level of engagement
ID Aim of the study Decisions modelled horizon Modelling approach Solution approach performance) with current practice
Rest and Mixed-integer programming Visit scheduling Single-period Mixed-integer Meta-heuristic approach based on Match of staff skills to patient Synthetic instances
Hirsch (2015) formulation and heuristic Allocation of staff (day) programming tabu search needs (Mismatch between with
algorithm for the problems Routing of visits allocated staff skills and realistic features
of scheduling, allocation and required skills)
routing in home health care. Punctuality (Tardiness)
Staff preferences (Deviation from
planned shift)
Staff travelling
(Travelling distance)
Rest and Mixed-integer programming Staff rostering Single-period Mixed-integer Combination of dynamic Match of staff skills to patient Synthetic instances
Hirsch (2016) formulation and heuristic Visit scheduling (day) programming programming (solving the needs (Mismatch between with
algorithm for the problems Allocation of staff shortest path problem between allocated staff skills and realistic features
of rostering, scheduling, Routing of visits locations at different times of required skills)
allocation and routing in Short-term fixes the day) and meta-heuristics Punctuality (Tardiness)
home health care. based on tabu search Staff preferences (Deviation from
planned shift)
Staff travelling (Travelling time)
Rodriguez Stochastic programming Team size and Single-period Two-stage, integer Branch-and-cut algorithm Staff costs (Wage costs) None
et al. (2015) approach for the problem of composition (several months) linear programming
staff dimensioning over a
time horizon in home
health care.
Shao Mixed-integer programming Visit scheduling Multi-period Mixed-integer Greedy randomised adaptive Costs of staff time spent with Synthetic instances
et al. (2012) formulation and heuristic Allocation of staff (days) programming search procedure patient (Treatment costs) with
algorithm for the problems Routing of visits Overtime (Overtime costs) realistic features
of scheduling, allocation and Staff productivity (Ratio between
routing in home health care. treatment cost and total cost)
Staff travelling (Travelling costs)
Shi Mixed-integer programming Visit scheduling Single-period Mixed-integer Hybrid genetic algorithm, with Staff travelling None
et al. (2017a) formulation and heuristic Allocation of staff (day) programming stochastic simulation to (Travelling distance)
algorithm for the problems Routing of visits evaluate candidate solutions
of scheduling, allocation and
routing in home health care,
with uncertainty in capacity
constraints modelled using
fuzzy credibility theory.
Shi Stochastic programming Visit scheduling Single-period Stochastic Hybrid genetic algorithm Patient preferences (Deviation None
et al. (2017b) formulation and heuristic Allocation of staff (day) programming from preferred visit time)
algorithm for the problems Routing of visits Staff preferences (Overtime
of scheduling, allocation and costs)
routing in home health care. Staff travelling (Travelling costs)
Shi et al. (2018) Stochastic programming Visit scheduling Single-period Stochastic Heuristic algorithm based on Patient preferences (Deviation None
formulation and heuristic Allocation of staff (day) programming simulated annealing from preferred visit time)
algorithm for the problems Routing of visits Staff preferences (Overtime
JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY
from organisations
Instances obtained
patients, appointment schedules and the routing of
staff are decided in combination.
approximation method
patient groups
formulations to support
price algorithm for the
resource allocation to
Stochastic programming
patients in long-term
Integer programming
home health care.
et al. (2015)
Figure 3. The set of partial decision hierarchies of decision among the 59 of 77 reviewed papers that considered more than
one of the identified decision problems in home health care. Each tile shows a different subset of decisions, with the number
of papers considering precisely that combination of decisions given at the top right of the tile. The line-style used for the
boxes denotes whether, within those papers, the respective decisions are: merely mentioned as relevant; viewed as a neces-
sary prior step but not modelled; modelled at the first (or only) step of analysis; or modelled at a second step of analysis.
JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY 1979
Table 2. Continued.
Decision (number of papers) OR approach (number of papers) Solution method (reference)
Combination of tabu search and adaptive large neighbourhood search
(Nickel et al., 2012)
Stochastic programming (n ¼ 7) Analytical derivation of optimal solution (Lanzarone & Matta, 2012)
Genetic algorithm (Shi et al., 2017b)
Simulated annealing (Shi et al., 2018)
Branch and price (Yuan et al., 2015; Yuan et al., 2018)
Not specified - commercially available software (Lanzarone et al., 2012;
Errarhout et al., 2016)
Cardinality-constrained Other heuristic algorithm (Cappanera et al., 2018)
approach (n ¼ 2) Not specified - commercially available software (Carello &
Lanzarone, 2014)
Multi-agent modelling (n ¼ 1) Multi-agent simulation (L opez-Santana et al., 2016)
Clustering (n ¼ 1) Greedy algorithm (Quintana et al., 2017)
Visit scheduling (n ¼ 54) (Mixed-)integer Insertion heuristics (Bennett & Erera, 2011)
programming (n ¼ 43) Savings algorithm (Bowers et al., 2015)
Particle swarm optimisation algorithm (Akjiratikarl et al., 2007)
Greedy randomised adaptive search procedure (Shao et al., 2012; Bard,
Shao & Jarrah, 2014)
Branch price and cut (Bard, Shao, Qi, et al. 2014; Trautsamwieser &
Hirsch, 2014)
Large neighbourhood search (Braekers et al., 2016)
Adaptive large neighbourhood search (Guericke & Suhl, 2017)
Branch and bound (Cappanera & Scutella, 2015)
Memetic algorithm (Decerle et al., 2018)
Genetic algorithm (Liu et al., 2013; Mutingi & Mbohwa, 2013; Du et al.,
2017; Shi et al., 2017a)
Combination of clustering and variable neighbourhood search (Erdem &
Bulkan, 2017)
Repeated matching (Eveborn et al., 2006; Eveborn et al., 2009)
Tabu search (Liu et al., 2013; Liu et al., 2014; Fikar & Hirsch, 2015; Rest &
Hirsch, 2015)
Discrete event driven metaheuristics (Fikar et al., 2016)
Record-to-record travel algorithm (Hewitt et al., 2016)
Harmony search (Lin et al., 2018)
Branch and price (Rasmussen et al., 2012; Manerba & Mansini, 2016; Liu
et al., 2017)
Fuzzy-based particle swarm optimisation algorithm (Mutingi & Mbohwa,
2014)
Variable neighbourhood search (Trautsamwieser et al., 2011;
Trautsamwieser & Hirsch, 2011; Nasir & Dang, 2018)
Combination of dynamic programming and tabu search (Rest & Hirsch,
2016)
Other heuristic algorithm (An et al., 2012; Bard, Shao, Qi, et al., 2014;
Mankowska et al., 2014; Maya Duque et al., 2015; Decerle et al., 2016;
Redjem & Marcon, 2016; Liu et al., 2018)
Not specified - commercially available software (Bachouch et al., 2011;
En-Nahli et al., 2015; Nasir & Dang, 2016; Yalçındag et al. 2016; Nasir
et al., 2018)
Constraint programming (n ¼ 4) Combination of tabu search and simulated annealing (Bertels & Fahle,
2006)
Large neighbourhood search (Cattafi et al., 2015)
Variable neighbourhood search (Hiermann et al., 2015)
Memetic algorithm (Hiermann et al., 2015)
Scatter search (Hiermann et al., 2015)
Simulated annealing (Hiermann et al., 2015)
Combination of tabu search and adaptive large neighbourhood search
(Nickel et al., 2012)
Cardinality-constrained Other heuristic algorithm (Cappanera et al., 2018)
approach (n ¼ 1)
Multi-agent modelling (n ¼ 1) Multi-agent simulation (Lopez-Santana et al., 2016)
Clustering (n ¼ 1) Greedy algorithm (Quintana et al., 2017)
Stochastic programming (n ¼ 4) Genetic algorithm (Shi et al., 2017b)
Simulated annealing (Shi et al., 2018)
Branch and price (Yuan et al., 2015; Yuan et al., 2018)
Routing of visits (n ¼ 55) (Mixed-)integer Savings algorithm and insertion heuristics (Begur et al., 1997)
programming (n ¼ 44) Insertion heuristics (Bennett & Erera, 2011)
Savings algorithm (Bowers et al., 2015)
Particle swarm optimisation algorithm (Akjiratikarl et al., 2007)
Greedy randomised adaptive search procedure (Shao et al., 2012; Bard,
Shao & Jarrah, 2014)
Branch price and cut (Bard, Shao, Qi, et al., 2014; Trautsamwieser &
Hirsch, 2014)
Large neighbourhood search (Braekers et al., 2016)
Adaptive large neighbourhood search (Guericke & Suhl, 2017)
Branch and bound (Cappanera & Scutella, 2015)
Memetic algorithm (Decerle et al., 2018)
Genetic algorithm (Liu et al., 2013; Mutingi & Mbohwa, 2013; Du et al.,
(continued)
JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY 1981
Table 2. Continued.
Decision (number of papers) OR approach (number of papers) Solution method (reference)
2017; Shi et al., 2017a)
Combination of clustering and variable neighbourhood search (Erdem &
Bulkan, 2017)
Repeated matching (Eveborn et al., 2006; Eveborn et al., 2009)
Tabu search (Liu et al., 2013; Liu et al., 2014; Fikar & Hirsch, 2015; Rest &
Hirsch, 2015)
Discrete event driven metaheuristics (Fikar et al., 2016)
Record-to-record travel algorithm (Hewitt et al., 2016)
Harmony search (Lin et al., 2018)
Branch and price (Rasmussen et al., 2012; Manerba & Mansini, 2016; Liu
et al., 2017)
Fuzzy-based particle swarm optimisation algorithm (Mutingi & Mbohwa,
2014)
Variable neighbourhood search (Trautsamwieser et al., 2011;
Trautsamwieser & Hirsch, 2011; Nasir & Dang, 2018)
Combination of dynamic programming and tabu search (Rest & Hirsch,
2016)
Other heuristic algorithm (An et al., 2012; Bard et al., 2013; Bard, Shao,
Qi, et al., 2014; Mankowska et al., 2014; Maya Duque et al., 2015;
Decerle et al., 2016; Redjem & Marcon, 2016; Liu et al., 2018)
Not specified - commercially available software (Bachouch et al., 2011;
Bard et al., 2013; En-Nahli et al., 2015; Nasir & Dang, 2016; Yalçındag
et al. 2016)
Constraint programming (n ¼ 4) Combination of tabu search and simulated annealing (Bertels & Fahle,
2006)
Large neighbourhood search (Cattafi et al., 2015)
Variable neighbourhood search (Hiermann et al., 2015)
Memetic algorithm (Hiermann et al., 2015)
Scatter search (Hiermann et al., 2015)
Simulated annealing (Hiermann et al., 2015)
Combination of tabu search and adaptive large neighbourhood search
(Nickel et al., 2012)
Cardinality-constrained Other heuristic algorithm (Cappanera et al., 2018)
approach (n ¼ 1)
Multi-agent modelling (n ¼ 1) Multi-agent simulation (Lopez-Santana et al., 2016)
Clustering (n ¼ 1) Greedy algorithm (Quintana et al., 2017)
Stochastic programming (n ¼ 4) Genetic algorithm (Shi et al., 2017b)
Simulated annealing (Shi et al., 2018)
Branch and price (Yuan et al., 2015; Yuan et al., 2018)
Coordination with other (Mixed-)integer Memetic algorithm (Decerle et al., 2018)
home care programming (n ¼ 4) Other heuristic algorithm (Decerle et al., 2016; Mankowska et al., 2014)
services (n ¼ 4) Not specified - commercially available software (Nasir et al., 2018)
Short-term fixes (n ¼ 9) (Mixed-)integer Savings algorithm and insertion heuristics (Begur et al., 1997)
programming (n ¼ 6) Repeated matching (Eveborn et al., 2006; Eveborn et al., 2009)
Harmony search (Lin et al., 2018)
Combination of dynamic programming and tabu search (Rest & Hirsch,
2016)
Tabu search (Yuan & Jiang, 2017)
Multi-agent modelling (n ¼ 2) Multi-agent simulation (Lopez-Santana et al., 2016; Marcon et al., 2017)
Constraint programming (n ¼ 1) Combination of tabu search and adaptive large neighbourhood search
(Nickel et al., 2012)
Reassessment (n ¼ 0) (n ¼ 0) –
Nahli et al., 2015), constraint programming (Nickel Coordination of care between different professio-
et al., 2012; Cattafi et al., 2015), and clustering techni- nals or different services is modelled using (mixed)-
ques (Quintana et al., 2017). Short-term fixes to such integer programming approaches and solved using
solutions are also addressed in several of these publica- bespoke heuristic algorithms (Mankowska et al.,
tions. For instance, Lin et al. (2018) use mixed-integer 2014; Decerle et al., 2016; Decerle et al., 2018).
programming and Nickel et al. (2012) use constraint
programming, both proposing two-step approaches: the 3.2.4. Approaches from the non-HHC literature
first step determining a rostering/scheduling/allocation/ In this section, we summarise our findings on OR
routing solution, the second step minimising deviations approaches to analogous decisions in settings other
from that solution in case of disruptions. Focused on than HHC. In particular, we focus on decisions that
practice, Begur et al. (1997), Eveborn et al. (2006) and are not currently covered by the HHC literature or
Eveborn et al. (2009) also use (mixed-)integer program- that are covered but might benefit from OR
ming as a core component of interactive decision sup- approaches adopted in other fields.
port tools producing solutions for a scheduling/ The problem of service definition and access has
allocation/routing problem and enabling the user to similarities to investment portfolio selection: choos-
introduce real-time barriers downstream, with solutions ing the set of home health care services to offer and
automatically adapted. identifying the corresponding target patients can be
1982 L. GRIECO ET AL.
seen as a form of investment selection (with return Among the cited references, Barua et al. (1997)
expressed as a combination of monetary and utility develop a game theory model of information exchange
terms) over a given planning period. Masmoudi to analyse the exchange of usable intra- and inter-
et al. (2018) review multiple objective deterministic organisational information between decision units and
and stochastic programming models for this prob- use the idea of “cheap talk” as a mechanism for coor-
lem. The aim is to select the least risky set of invest- dinating inter-organisational activities.
ments (services in HHC settings) for a given level of In our search we did not identify decisions dir-
return, which need not be monetary. As discussed ectly analogous to the design of processes for
in deLlano-Paz et al. (2017), the recent use of port- reassessment of patient need. However, we consider
folio theory in energy planning challenges least-cost that there are similarities to optimal control prob-
approaches that favour fossil fuels and can incorpor- lems, consisting of designing and dimensioning
ate multiple stakeholder perspectives. interventions on a system over time (and possibly
While potentially unpalatable to many working in accounting for measurement uncertainties) in order
HHC, time-of-use pricing could have useful insights to optimise the behaviour of that system. In HHC
for patient screening. For instance, Chrysopoulos settings, such interventions would consist of updat-
and Mitkas (2018) discuss the use of multi-objective ing patients’ home health care plans based on
particle swarm optimisation to customise time-of- changes to their health status. For instance,
use pricing to deter some customers from requesting B€uy€uktahtakın and Haight (2018) reviewed such an
service at peak times. approach in the markedly different context of inva-
Related to team size and composition, Boucher
sive species management.
et al. (2007) survey how competence-related con- Finally, our search did not identify OR literature
cepts are modelled and formalised in the literature
related to decisions analogous to commissioning or
on industrial performance improvement. This litera-
workforce role definition, nor work outside HHC to
ture focuses on interaction and complementarity
augment approaches to the decisions of allocation,
between team members, distinct from the HHC lit-
scheduling, routing or short-term fixes.
erature focus on just the team configurations that
best meet demand. Aspects analysed include team
3.2.5. Aspects of system performance in home
building and competence allocation. Specifically,
health care
studies are cited that start from “Belbin’s self-per-
As part of the data extraction, we identified from
ception inventory” to develop formal models and/or
the stated aims of each paper, or from the objective
decision support systems to assist managers in team
functions and constraints within the models pro-
composition (e.g. Sommerville & Dalziel, 1998;
posed, the different aspects of the HHC system’s
Chen & Lin, 2004).
performance considered to contribute to service
Decisions on the home health care plans made for
quality in that paper (for instance workload balance
patients can be seen as analogous to the elicitation of
and response to “customer requirements”. A poten- or continuity of care).
We identified 21 different aspects of system per-
tially relevant stream of research is that dealing with
intelligent design for customised products, defined by formance and 55 different related metrics (see
Zhang et al. (2017) as products “designed to satisfy Appendix 2) across the 77 accepted articles. Table 3
the customer’s individual and diversified requirements reports the number of papers that consider each
as quickly and at as low a cost as possible”. They cite aspect for each decision problem (see Figure 1 for a
Quality Function Deployment (Akao, 2004) as one definition of the decisions).
approach to map product features to customer We briefly discuss below the aspects of system
requirements that has been used in service design (see performance that authors use to define the quality
for instance Yan et al., 2016). of solutions to different decision problems and their
For the decision of staff rostering, the HHC lit- alignment with the perspectives of different stake-
erature found in our systematic review does not holders in HHC. First, it is worth noting that over
address shift design, whereas in other contexts this 30 of the papers reviewed explicitly consider the
problem is addressed using queueing theory (Green productivity of the staff responsible for making
et al., 2006) and integer programming (Ingolfsson these decisions. This acknowledgement of the (often
et al., 2010). considerable) planning time that staff currently
Approaches analysing the benefits of information devote to these tasks and the extent to which this
flow within and between organisations seem relevant could be reduced by the adoption of OR solutions
to the coordination of home care between services. brings a focus in the literature on the process of
Durugbo et al. (2013) review quantitative flow analysis decision support and on heuristic approaches that
methods for improving organisational performance. reduce the high computational times associated with
JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY 1983
existing solution methods (e.g. An et al., 2012; We found few papers on the commissioning of
Eveborn et al., 2006). HHC and the commissioner’s perspective is largely
Due to the dominance in the literature of sched- absent, although many models penalise or preclude
uling, allocation and routing problems (usually solutions that fail to meet demand and one paper
modelled as a vehicle routing problem), staff travel- seeks to maximise the number of new patients
ling is the most represented aspect of system per- accepted per week (De Angelis, 1998).
formance. In particular, a common objective is to As with most complex systems, the models devel-
minimise travelling costs (or distance, or time) oped in HHC reflect intrinsic tensions and trade-offs
incurred by the provider for daily home visits. between different aspects of system performance. For
While this reflects a provider perspective (cost mini- instance, Benzarti et al. (2013) built two complemen-
misation), some authors augment their formulation tary models to solve the districting problem: i) one
to penalise or preclude solutions that do not take minimising the maximum size of districts (seen as
the presumed or elicited interests of patients and/or desirable), with constraints on the imbalance of work-
staff into account. load between districts; ii) one minimising the max-
Patient perspectives are manifest in preferred imum deviation of district workload from the average
time slots for visits (“time windows”) in the model workload, with maximum district size constrained.
formulation, either as hard constraints (Du et al., Their results show a clear tension between the two
2017; Akjiratikarl et al., 2007), or penalties in the aspects. The more common approach of combining
objective function (Braekers et al., 2016; Mankowska contrasting aspects of system performance in a single
et al., 2014), or both (Bertels & Fahle, 2006). Overall objective function is exemplified in Nasir and Dang
tardiness is also used as a measure of patient experi- (2018), who tackle the decision of team size and com-
ence and modelled as a component of the objective position by minimising a weighted sum of: staff travel-
function to be minimised (Mankowska et al., 2014; ling (measured through total distance travelled), staff
costs (measured through wage costs), workload bal-
Rest & Hirsch 2016). Some papers consider patients’
ance (measured through workload variability across
preference for specific staff to carry out the visit
staff), and demand satisfaction (measured through a
(usually through penalties in the objective function)
cost of delaying patients).
(Bowers et al., 2015; Braekers et al., 2016).
While only a handful of papers study decisions
Continuity of care is another aspect of system
that span different levels of planning, some aspects
performance important from a patient perspective.
of system performance are considered in strategic,
Ensuring that the same staff visit a given patient
tactical and operational models. Specifically, staff
during the whole period of care can be in conflict
travelling is considered in decisions of districting
with workload balance (staff perspective) and cost
(Blais et al., 2003), team size and composition
containment (provider perspective), however it has
(Nasir & Dang, 2018), screening (Nasir & Dang,
been shown that OR approaches can achieve a good
2018) and in the vast majority of papers on roster-
level of continuity of care and a fair workload bal-
ing/scheduling/allocation/routing. In all these stud-
ance with limited cost increase (Carello et al., 2018). ies, the total distance travelled by staff is used
The most common measure for continuity of care is consistently as a metric.
the number of staff per patient (Cattafi et al., 2015; The level of demand satisfied is considered in
Hewitt et al., 2016), though other metrics are also models addressing decisions of screening and team
used such as number of staff deployed on tours of size and composition (Nasir & Dang, 2018) and
similar characteristics (Wirnitzer et al., 2016) or joint decisions of rostering, scheduling, allocation
number of changes in the allocation of staff to a and routing (Nickel et al., 2012), using penalty costs
patient (Wirnitzer et al., 2016; Carello et al., 2018). for patient delays and unassigned visits respectively.
A related aspect of system performance that Workload balance among teams or individuals
reflects a staff perspective is workload balance, with also features at different planning levels, with
this measured in several publications through met- Benzarti et al. (2013) minimising the maximum
rics such as the deviation of each member of staff’s deviation of district workload from the average in
workload from average workload (e.g. Benzarti strategic districting decisions and Cappanera et al.
et al., 2013). Some models account directly for a (2015) maximising the minimum staff utilisation
staff perspective by incorporating preferences for rate within a district in operational decisions of
certain patients (Hiermann et al., 2015) or penalisa- scheduling, allocation and routing.
tion of staff waiting time (Trautsamwieser et al., While these single aspects of system performance
2011). Also, we note that approaches that minimise appear in papers across planning levels, clustering
staff travel time or overtime (Braekers et al., 2016) papers by the sets of performance measures consid-
may also reflect the perspectives of some staff. ered tended to group papers addressing similar
1984 L. GRIECO ET AL.
Synchronisation Workload
decisions respectively. For each pair {decision, aspect of performance}, the table reports the number of papers modelling that decision while mentioning that aspect of performance. For instance, the aspect of performance
“Overtime” appears in 10 papers: 3 dealing with staff rostering, 10 dealing with allocation of staff, 8 dealing with routing of visits, 1 dealing with short-term fixes (note that there are papers dealing with more than one decision
In the column and row names, we indicate (in brackets) the number of papers featuring each aspect of performance and each decision, respectively. See Appendix 2 and Figure 1 for definitions of the aspects of performance and
(n ¼ 14)
balance
11
2
1
6
6
ters of papers identified (see Appendix 3), two
spanned planning levels, perhaps the most useful
of visits
(n ¼ 1)
being a cluster of 9 papers that use a similar set of
1
1
1
1
performance measures in dealing with decisions on
districting and (in different combinations) rostering,
Punctuality
(n ¼ 5)
1
4
4
4
1
1
Disruptions
Demand to planned
(n ¼ 3)
3
Part of our motivation for this review was to inform
sufficiency size autonomy preferences with patient costs productivity utilisation productivity travelling Overtime satisfaction
2
9
8
8
1
the delivery of HHC, and so we had a particular
interest in how the home care workforce was con-
(n ¼ 50) (n ¼ 10)
10
45
45
45
2
1
(n ¼ 31)
1
5
5
5
1
5
6
7
1
3
3
2
1
(n ¼ 8)
1
8
7
8
1
13
13
13
3
(n ¼ 1)
organisation.
skills to
(n ¼ 7)
1
7
7
6
1
22
22
21
1
8
8
care services (n ¼ 4)
Short-term fixes (n ¼ 9)
Workforce roles (n ¼ 0)
Reassessment (n ¼ 0)
Screening (n ¼ 4)
reflecting a view in the organisation studied that Both works by Hindle et al. (2000; 2009) were
this would benefit all parties. conceived to provide health care organisations with
For systems where nurses are considered perman- inputs for decisions on fund allocation (commis-
ent salaried staff, while there is some prospect of sioning) at national level. The former study
cost reduction through reduced overtime or optimal informed yearly allocation of funds for HHC to four
dimensioning of the workforce, the objectives of areas in Northern Ireland, improving existing crite-
allocation, scheduling and routing models (e.g. ria already based on OR modelling by incorporating
Carello et al., 2018; Decerle et al., 2018; L opez- a “rurality” feature in the model. The latter, aimed
Santana et al., 2016) focus more on the opportunity to support HHC funding allocation across districts
costs of travel times and unutilised staff time. in England, was less successful due to the low sur-
Workload balance within or across teams also vey response rate by local authorities in providing
features in some papers (for instance Cattafi et al., key input data to the researchers.
2015; Bowers et al., 2015). In the main, this is con- Begur et al. (1997), Eveborn et al. (2006) and
sidered simply in terms of the number of visits allo- Eveborn et al. (2009) discuss the use of (mixed-)integer
cated to staff, but Hertz and Lahrichi (2009) assess programming as a core component of interactive deci-
instead the burden that different types of case pre- sion support tools producing solutions for the schedul-
sent for different grades of staff. ing/allocation/routing problem and enabling the user
Another feature of modelled systems is the degree to introduce real-time barriers downstream, with solu-
of autonomy or agency that staff have. A number of tions automatically adapted (short-term fixes). In
modelled systems have scope for staff to express another scheduling/allocation/routing example, a HHC
preferences over when they work (Maya Duque organisation implemented and tested the approach pro-
et al., 2015; Bertels & Fahle, 2006). In the setting posed by Maya Duque et al. (2015), in collaboration
studied by Hertz and Lahrichi (2009), nurses have with an external software company. The authors
full autonomy over how they arrange their working assisted the organisation during the early stage of
day, with this precluding the development and use implementation by developing and testing a prototype
of centralised routing and scheduling algorithms. In of the tool. They also worked with a district manager
other works, the modelled systems operate such that to compare solutions from the tool with manually-gen-
optimisation could work to the detriment of staff, erated ones and to understand the level of acceptability
for instance where staff travel costs and the time to of the former. The organisation also provided advice
their first appointment and home from their last during the modelling process that helped to shape the
appointment are not included in the objective func- approach taken. For instance, the authors present a
tion (e.g. Bennett & Erera, 2011) or where solutions two-stage solution strategy not relying on Pareto-opti-
are in line with the letter but perhaps not the spirit mality criteria because the organisation considered pref-
of working regulations (the 7-hour shifts in Cattafi erences to be more important than total distance
et al., 2015 are partly driven by regulation that travelled. Finally, software implementing the staff allo-
workers should have a half-hour break every 7 h). cation models developed by Lanzarone et al. (2012)
was adopted by a HHC provider.
3.2.7. Methods vs practice More often, algorithms in the literature were tested
Most of the papers reviewed focus on the develop- on simulated instances or instances generated from
ment of new formulations and/or fast solution algo- data provided by partner organisations. Among the
rithms to better deal with model complexity. Few latter, the model presented by Trautsamwieser et al.
studies address the challenges of implementing the (2011) for rostering, scheduling, allocation and routing
proposed approach in practice and the implications decisions was developed in close connection with a
that these challenges have for the development of health care organisation with plans to include it in
models and solution methods. We highlight some their decision support systems; Wirnitzer et al. (2016)
exceptions in what follows. mention a partnership with a home care organisation
Although using randomly generated instances, who provided a test instance and contributed to the
Benzarti et al. (2013) present their work on district- assessment of different objective functions, one of
ing as a decision support system for managers in which was used (manually) by the organisation; close
HHC, envisaging that staff would interact with the collaboration with health care organisations who
tool (by modifying relevant parameters) and adapt actively contributed to the modelling work is also
the suggested solutions. The districting solution mentioned in Cattafi et al. (2015), although no details
obtained by Blais et al. (2003) was actually adopted were given on any implementation. Among the papers
by their client who had provided the authors with related to patient screening, only Zhang et al. (2015)
real-world instances to validate the model. mention a collaboration with an HHC organisation,
1986 L. GRIECO ET AL.
which in their case provided advice on current admis- We found that the OR literature concerning
sion policies. HHC remains dominated by papers proposing mod-
els and solution methods for combinations of oper-
ational decisions concerning the rostering of staff,
4. Discussion
the allocation of staff to patient visits, the schedul-
The strengths and weaknesses of a systematic ing of visits and the routing of staff. Previous
approach to reviewing the literature are well rehearsed; reviews (Cisse et al., 2017 and Fikar & Hirsch, 2017)
while a reproducible and transparent approach to have suggested a need for more stochastic formula-
identifying papers for review, we acknowledge that tions in such models and our review found that
search and selection of papers driven by keywords recent work has indeed incorporated stochastic
and prospective criteria can fail to identify relevant patient demand, staff travel times and service times
papers that could be found through other approaches. (for instance Shi et al., 2017a, 2017b, 2018; Yuan
Although this has been mitigated by considering the et al., 2018; Cappanera et al., 2018). Other advances
reference lists of previous (non-systematic) review since previous reviews include moves to explore
articles, we do not claim that our review contains all alternative approaches to the multiple aspects of sys-
relevant papers. Additionally, we note that our deci- tem performance acknowledged to be important in
sion framework is not exhaustive and that individuals, HHC, with for instance Liu et al. (2018) and Carello
teams and organisations may face other decision prob- et al. (2018) using multi-objective optimisation, as
lems specific to home health care that are amenable to suggested by Cisse et al. (2017). This ambition to
Operational Research approaches. The other limitation address more of the complexity inherent in HHC
intrinsic to our study design is that there is an inevit- systems motivates the continued focus within the
able degree of subjectivity in some of our interpret- literature on the development of computationally
ation of the intent and context of papers in the efficient solution methods.
literature. That said, this systematic review and ana- Our analysis shows some recognition of the hier-
lysis of the OR literature on decision problems in archy of decisions in home health care with many
home health care has yielded some valuable findings1. partial hierarchies addressed, but as stated above it
Previous reviews indicated that individual papers remains the case that papers addressing strategic,
did not span the strategic, tactical and operational tactical and operational decisions are missing from
decisions faced in home health care. By studying the this literature. One defence of the research commu-
combinations and hierarchies of decision addressed nity on this point would be that Operational
in individual papers and by clustering papers that Researchers often take the pragmatic view of want-
have a coherent view on system performance, one ing to build models for the problem in front of
contribution of this work is to establish that there them, which typically means models that serve the
are no groups of papers that, between them, provide explicit decisions made by the organisation or team
a comprehensive and coherent set of tools for stra- within an organisation that they are working with
tegic, tactical and operational decision-making in and, to a greater or lesser extent, adopting the per-
home health care. So while a home health care spective of that organisation or team. Another prag-
organisation could find approaches in the literature matic consideration would be that strategic decision
to problems at different planning levels, there would making often has a periodicity of many years with
be no guarantee that tools chosen for operational organisations rarely receptive to solutions that
decisions would not undermine the value of tools involve structural change unless they arrive at the
chosen for tactical or strategic decisions, due to the right point in this cycle, whatever the prom-
disparate aspects of system performance considered ised benefits.
in different papers. Despite the negative results in This sets up a tension between the potential value
this instance, as far as we are aware this use of clus- of “whole supply-chain” approaches and pragmatic
tering is novel in the context of systematic reviewing considerations of implementation. This would be a
and has potential to enhance analyses of other stronger defence if there was a greater focus on
literatures. implementation in this literature, yet few papers
Other contributions of our work are that it has report successful implementation in the organisa-
identified decisions in home health care amenable tions they worked with or that supplied data.
to operational research where there has been little Indeed, there is arguably little value to home health
or no research activity, and that is has highlighted care in further improvement of models and solution
contextual and environmental factors that should methods without greater attention given to chal-
perhaps be included to enhance OR papers on lenges to implementation, and we repeat the call of
home health care problems. These are pre- Fikar and Hirsch (2017) for more study of the
sented below. organisational and social factors that inhibit and
JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY 1987
promote the adoption of innovations rooted in OR configurations of HHC service and in what societal
to decision processes in HHC. settings a model could be applied or adapted.
Our review also highlights strategic and tactical Based on this review, we make a number of rec-
decisions where there is insufficient literature. ommendations for future research. We consider that
Specifically, we found little on the coordination of work addressing problems of role definition, home
care across professions and organisations (although health care plan construction and contract design
Nasir et al., 2018 make a valuable start on this), the for commissioning would be valuable additions to
design of contracts to support effective commission- the literature, as would models to support the
ing in HHC, the construction of home health care coordination and integration of different home
plans and role definitions within the workforce. based health and care services. In conducting and
The lack of research on role definition is indica- reporting such work, we suggest that it would be
tive of researchers accepting some of the choices beneficial for authors to align, where possible, the
made in HHC as fixed, or rather as not being aspects of system performance used in objectives
choices. This fits a general pattern in the literature and constraints with those used in existing litera-
of HHC systems being presented as characterising ture. This would enhance the prospects of decision
“the HHC problem” rather than one instance of a makers identifying a suite of OR models that
HHC problem specific to the system or organisation address decisions at different planning levels but
that informed model development. For instance, that promote consistent goals. Also, as discussed
models were, naturally, strongly influenced by the above, we consider that it would be useful for
employment practices of the organisations studied authors to clarify the context of the home health
and labour protections in place in that setting. care organisations that they work with such that the
Other aspects of organisational context are appar- applicability of models to different societal settings
ent from, though not always made explicit in, can be more readily assessed.
descriptions of the organisational problems faced
and in the models developed to address these prob-
Note
lems. These include features of the environment in
which an organisation is operating, such as whether 1. Given the elapsed time between running our
it can choose what home care services it offers, the literature search and publication of this article, we re-
ran our search during the peer-review process (on the
profiles of patient groups it accommodates or the 20th January 2020) and obtained 63 additional
individual patients it takes on within these groups, articles, of which 34 were retained based on title/
and what it is paid to do and on what terms. In abstract. Based on information gathered from the
some settings, these aspects constitute a choice of abstracts and full-text, we found that all of the
business model, in other settings they reflect the identified papers reported formulations and/or
entitlements of citizens and the organisation of state algorithms for some combination of allocation,
scheduling and routing problems, except one paper
provision or state funded provision. Similarly, the presenting a formulation and an algorithm for the
overarching organisational objective may be profit- districting problem. We did not identify any relevant
ability, sustainability, meeting an explicit or implicit discrepancies between such additional results and the
social contract, improving population health, or pre- findings reported in this review.
venting or delaying the use of more expensive
health services. Disclosure statement
Other features often excluded from problem
descriptions include the proportion of the organisa- No potential conflict of interest was reported by the author(s).
tion’s business that is represented by HHC, or spe-
cifically whether HHC is organised independently Funding
from its other services. This is relevant when con-
This work was funded by The Health Foundation, an
sidering the value of a whole supply-chain approach independent charity working to continuously improve the
because, if strategic decisions are made for a wider quality of healthcare in the UK (award reference number:
set of services, then good strategy may not necessar- 79808). The funders had no role in study design, data col-
ily be conducive to good operational performance in lection and analysis, decision to publish, or preparation of
HHC services. the manuscript.
While the omission of these features of organisa-
tional context does not necessarily undermine the
potential value of the work done for the organisa- ORCID
tions studied, it can be argued that it makes the Luca Grieco http://orcid.org/0000-0002-8733-6062
models less generalisable (Sahin & Matta, 2015), or Martin Utley http://orcid.org/0000-0001-9928-1516
rather it makes it harder to understand what Sonya Crowe http://orcid.org/0000-0003-1882-5476
1988 L. GRIECO ET AL.
deLlano-Paz, F., Calvo-Silvosa, A., Antelo, S. I., & Soares, Hewitt, M., Nowak, M., & Nataraj, N. (2016). Planning
I. (2017). Energy planning and modern portfolio the- strategies for home health care delivery. Asia-Pacific
ory: A review. Renewable and Sustainable Energy Journal of Operational Research, 33(05), 1650041.
Reviews, 77, 636–651. https://doi.org/10.1016/j.rser. https://doi.org/10.1142/S021759591650041X
2017.04.045 Hiermann, G., Prandtstetter, M., Rendl, A., Puchinger, J.,
Du, G., Liang, X., & Sun, C. (2017). Scheduling optimiza- & Raidl, G. R. (2015). Metaheuristics for solving a
tion of home health care service considering patients’ multimodal home-healthcare scheduling problem.
priorities and time windows. Sustainability, 9(2), 253. Central European Journal of Operations Research, 23(1),
https://doi.org/10.3390/su9020253 89–113. https://doi.org/10.1007/s10100-013-0305-8
Durugbo, C., Tiwari, A., & Alcock, J. R. (2013). Hindle, T., Hindle, A., & Spollen, M. (2000). Resource
Modelling information flow for organisations: A review allocation modelling for home-based health and social
of approaches and future challenges. International care services in areas having differential population
Journal of Information Management, 33(3), 597–610. density levels: A case study in Northern Ireland. Health
https://doi.org/10.1016/j.ijinfomgt.2013.01.009 Services Management Research, 13(3), 164–169. https://
En-Nahli, L., Allaoui, H., & Nouaouri, I. (2015). A multi- doi.org/10.1177/095148480001300304
objective modelling to human resource assignment and Hindle, T., Hindle, G., & Spollen, M. (2009). Travel-
routing problem for home health care services. IFAC- related costs of population dispersion in the provision
PapersOnLine, 48(3), 698–703. https://doi.org/10.1016/j. of domiciliary care to the elderly: A case study in
ifacol.2015.06.164 English Local Authorities. Health Services Management
Erdem, M., & Bulkan, S. (2017). A two-stage solution Research, 22(1), 27–32. https://doi.org/10.1258/hsmr.
approach for the large-scale home healthcare routeing 2008.008012
and scheduling problem. South African Journal of Hulshof, P. J. H., Kortbeek, N., Boucherie, R. J., Hans,
Industrial Engineering, 28(4), 133–149. https://doi.org/ E. W., & Bakker, P. J. M. (2012). Taxonomic classification
10.7166/28-4-1754 of planning decisions in health care: A structured review
Errarhout, A., Kharraja, S., & Corbier, C. (2016). Two- of the state of the art in OR/MS. Health Systems, 1(2),
stage stochastic assignment problem in the home health 129–175. https://doi.org/10.1057/hs.2012.18
care. IFAC-PapersOnLine, 49(12), 1152–1157. https:// Ingolfsson, A., Campello, F., Wu, X., & Cabral, E. (2010).
doi.org/10.1016/j.ifacol.2016.07.659 Combining integer programming and the randomiza-
Eveborn, P., Flisberg, P., & R€ onnqvist, M. (2006). Laps tion method to schedule employees. European Journal
of Operational Research, 202(1), 153–163. https://doi.
care—an operational system for staff planning of home
org/10.1016/j.ejor.2009.04.026
care. European Journal of Operational Research, 171(3),
Koeleman, P. M., Bhulai, S., & van Meersbergen, M.
962–976. https://doi.org/10.1016/j.ejor.2005.01.011
(2012). Optimal patient and personnel scheduling poli-
Eveborn, P., R€ onnqvist, M., Einarsdottir, H., Eklund, M.,
cies for care-at-home service facilities. European
Liden, K., & Almroth, M. (2009). Operations research
Journal of Operational Research, 219(3), 557–563.
improves quality and efficiency in home care.
https://doi.org/10.1016/j.ejor.2011.10.046
Interfaces, 39(1), 18–34. https://doi.org/10.1287/inte.
Lanzarone, E., & Matta, A. (2012). A cost assignment pol-
1080.0411
icy for home care patients. Flexible Services and
Fikar, C., & Hirsch, P. (2015). A matheuristic for routing
Manufacturing Journal, 24(4), 465–495. https://doi.org/
real-world home service transport systems facilitating 10.1007/s10696-011-9121-4
walking. Journal of Cleaner Production, 105, 300–310. Lanzarone, E., Matta, A., & Sahin, E. (2012). Operations
https://doi.org/10.1016/j.jclepro.2014.07.013 management applied to home care services: The prob-
Fikar, C., & Hirsch, P. (2017). Home health care routing lem of assigning human resources to patients. IEEE
and scheduling: A review. Computers & Operations Transactions on Systems, Man, and Cybernetics - Part
Research, 77, 86–95. https://doi.org/10.1016/j.cor.2016. A: Systems and Humans, 42(6), 1346–1363. https://doi.
07.019 org/10.1109/TSMCA.2012.2210207
Fikar, C., Juan, A. A., Martinez, E., & Hirsch, P. (2016). Lanzarone, E., Matta, A., & Scaccabarozzi, G. (2010). A
A discrete-event driven metaheuristic for dynamic patient stochastic model to support human resource
home service routing with synchronised trip sharing. planning in home care. Production Planning & Control,
European J. of Industrial Engineering, 10(3), 323. 21(1), 3–25. https://doi.org/10.1080/09537280903232362
https://doi.org/10.1504/EJIE.2016.076382 Lin, C.-C., Hung, L.-P., Liu, W.-Y., & Tsai, M.-C. (2018).
Green, L. V., Soares, J., Giglio, J. F., & Green, R. A. Jointly rostering, routing, and rerostering for home
(2006). Using queueing theory to increase the effective- health care services: A harmony search approach with
ness of emergency department provider staffing. genetic, saturation, inheritance, and immigrant
Academic Emergency Medicine, 13(1), 61–68. https:// schemes. Computers & Industrial Engineering, 115,
doi.org/10.1197/j.aem.2005.07.034 151–166. https://doi.org/10.1016/j.cie.2017.11.004
Guericke, D., & Suhl, L. (2017). The home health care Lin, M., Chin, K. S., Wang, X., & Tsui, K. L. (2016). The
problem with working regulations. Or Spectrum, 39(4), therapist assignment problem in home healthcare
977–1010. https://doi.org/10.1007/s00291-017-0481-3 structures. Expert Systems with Applications, 62, 44–62.
Gutierrez, E. V., & Vidal, C. J. (2013). Home health care https://doi.org/10.1016/j.eswa.2016.06.010
logistics management problems: A critical review of Lin, M., Chin, K.-S., Fu, C., & Tsui, K.-L. (2017). An effect-
models and methods. Revista Facultad de Ingenierıa ive greedy method for the meals-on-wheels service dis-
Universidad de Antioquia, 68, 160–175. tricting problem. Computers & Industrial Engineering,
Hertz, A., & Lahrichi, N. (2009). A patient assignment 106, 1–19. https://doi.org/10.1016/j.cie.2017.01.022
algorithm for home care services. Journal of the Liu, M., Yang, D., Su, Q., & Xu, L. (2018). Bi-objective
Operational Research Society, 60(4), 481–495. https:// approaches for home healthcare medical team planning
doi.org/10.1057/palgrave.jors.2602574 and scheduling problem. Computational and Applied
1990 L. GRIECO ET AL.
service time. IFAC-PapersOnLine, 50(1), 13987–13992. Yan, H.-B., Cai, S., & Li, M. (2016). A two-stage fuzzy
https://doi.org/10.1016/j.ifacol.2017.08.2419 quality function deployment model for service design.
Shi, Y., Boudouh, T., Grunder, O., & Wang, D. (2018). In V.-N. Huynh, M. Inuiguchi, B. Le, B. N. Le, & T.
Modeling and solving simultaneous delivery and pick- Denoeux (Eds.), Integrated uncertainty in knowledge
up problem with stochastic travel and service times in modelling and decision making. (Vol. 9978, pp.
home health care. Expert Systems with Applications, 110–123). Cham: Springer. https://doi.org/10.1007/978-
102, 218–233. https://doi.org/10.1016/j.eswa.2018.02.025 3-319-49046-5_10
Sommerville, J., & Dalziel, S. (1998). Project teambuild- Yuan, B., & Jiang, Z. (2017). Disruption management for
ing—the applicability of Belbin’s team-role self-percep- the real-time home caregiver scheduling and routing
tion inventory. International Journal of Project problem. Sustainability, 9(12), 2178. https://doi.org/10.
Management, 16(3), 165–171. https://doi.org/10.1016/ 3390/su9122178.
S0263-7863(97)00054-9 Yuan, B., Liu, R., & Jiang, Z. (2015). A branch-and-price
Trautsamwieser, A., Gronalt, M., & Hirsch, P. (2011). algorithm for the home health care scheduling and
Securing home health care in times of natural disasters. Or
routing problem with stochastic service times and skill
Spectrum, 33(3), 787–813. https://doi.org/10.1007/s00291-
requirements. International Journal of Production
011-0253-4
Research, 53(24), 7450–7464. https://doi.org/10.1080/
Trautsamwieser, A., & Hirsch, P. (2011). Optimization of
00207543.2015.1082041
daily scheduling for home healthcare services. Journal
Yuan, B., Liu, R., & Jiang, Z. (2018). Daily scheduling of
of Applied Operational Research, 3(3), 124–136.
Trautsamwieser, A., & Hirsch, P. (2014). A branch-price- caregivers with stochastic times. International Journal
and-cut approach for solving the medium-term home of Production Research, 56(9), 3245–3261. https://doi.
health care planning problem. Networks, 64(3), org/10.1080/00207543.2018.1445879
143–159. https://doi.org/10.1002/net.21566 Zhang, R., Fry, M. J., & Krishnan, H. (2015). Efficiency
Wirnitzer, J., Heckmann, I., Meyer, A., & Nickel, S. and Equity in Healthcare: An Analysis of Resource
(2016). Patient-based nurse rostering in home care. Allocation Decisions in a Long-Term Home Care
Operations Research for Health Care, 8, 91–102. https:// Setting. INFOR: Information Systems and Operational
doi.org/10.1016/j.orhc.2015.08.005 Research, 53(3), 97–112. https://doi.org/10.3138/infor.
Yalçındag, S., Cappanera, P., Grazia Scutella, M., Şahin, 53.3.97
E., & Matta, A. (2016). Pattern-based decompositions Zhang, S., Xu, J., Gou, H., & Tan, J. (2017). A research
for human resource planning in home health care serv- review on the key technologies of intelligent design for
ices. Computers & Operations Research, 73, 12–26. customized products. Engineering, 3(5), 631–640.
https://doi.org/10.1016/j.cor.2016.02.011 https://doi.org/10.1016/J.ENG.2017.04.005