ICU Early Physical Rehabilitation Programs: Financial Modeling of Cost Savings*

Robert K. Lord, AB1; Christopher R. Mayhew, BS1; Radha Korupolu, MBBS, MS2; Earl C. Mantheiy, BA1; Michael A. Friedman, PT, MBA3; Jeffrey B. Palmer, MD4; Dale M. Needham, FCA, MD, PhD1,4

Objective: To evaluate the potential annual net cost savings of implementing an ICU early rehabilitation program. Design: Using data from existing publications and actual experience with an early rehabilitation program in the Johns Hopkins Hospital Medical ICU, we developed a model of net nancial savings/costs and presented results for ICUs with 200, 600, 900, and 2,000 annual admissions, accounting for both conservativeand best-case scenarios. Our example scenario provided a projected nancial analysis of the Johns Hopkins Medical ICU early rehabilitation program, with 900 admissions per year, using actual reductions in length of stay achieved by this program. Setting: U.S.-based adult ICUs. Interventions: Financial modeling of the introduction of an ICU early rehabilitation program. Measurements and Main Results: Net cost savings generated in our example scenario, with 900 annual admissions and actual length of stay reductions of 22% and 19% for the ICU and oor, respectively, were $817,836. Sensitivity analyses, which used

conservative- and best-case scenarios for length of stay reductions and varied the per-day ICU and oor costs, across ICUs with 2002,000 annual admissions, yielded nancial projections ranging from $87,611 (net cost) to $3,763,149 (net savings). Of the 24 scenarios included in these sensitivity analyses, 20 (83%) demonstrated net savings, with a relatively small net cost occurring in the remaining four scenarios, mostly when simultaneously combining the most conservative assumptions. Conclusions: A nancial model, based on actual experience and published data, projects that investment in an ICU early rehabilitation program can generate net nancial savings for U.S. hospitals. Even under the most conservative assumptions, the projected net cost of implementing such a program is modest relative to the substantial improvements in patient outcomes demonstrated by ICU early rehabilitation programs. (Crit Care Med 2013; 41:717724) Key Words: cost savings; critical illness; early ambulation; hospital costs; ICUs; length of stay; mobility; physical therapy; rehabilitation

urvivors of critical illness frequently have important impairments in their physical function and quality of life (15). Recent studies evaluating the early introduction of physical rehabilitation in the ICU have demonstrated improvements in physical function and quality of life, and in posthospital readmissions, institutionalization, and mortality, as well as reductions in mechanical ventilation duration and ICU and hospital length of stay (LOS) (612).
*See also p. 909. 1 Outcomes After Critical Illness and Surgery (OACIS) Group, Division of Pulmonary and Critical Care Medicine, The Johns Hopkins University School of Medicine, Baltimore, MD. 2 Department of Physical Medicine and Rehabilitation, The University of  Kentucky College of Medicine, Lexington, KY. 3 Department of Physical Medicine and Rehabilitation, The Johns Hopkins  Hospital, Baltimore, MD. 4 Department of Physical Medicine and Rehabilitation, The Johns Hopkins  School of Medicine, Baltimore, MD. The authors have not disclosed any potential conicts of interest. For information regarding this article, E-mail: dale.needham@jhmi.edu Copyright 2013 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI: 10.1097/CCM.0b013e3182711de2

Despite these benets, the nancial impact of implementing new programs is a potential barrier to adoption (6, 13) since early rehabilitation programs may require additional staff and/or equipment (14). Existing studies have frequently demonstrated a decreased LOS due to early rehabilitation, with suggestions that such programs may be cost-neutral or result in net cost savings (6, 9, 10); however, no formal nancial analysis of the impact of such cost reductions has been published. In order to help with hospital-level decision making regarding the implementation of an ICU early rehabilitation program, our objective is to create a nancial model capable of analyzing the net savings or costs associated with a U.S. hospital implementing an ICU early rehabilitation program.

METHODS
Financial Model Development Using a methodology similar to a prior ICU nancial analysis evaluating another ICU intervention (13), a model was designed to assess the net nancial savings or costs for a
www.ccmjournal.org

Critical Care Medicine

717

Lord et al

hospital implementing an ICU early rehabilitation program. The model considered the following key variables: 1) the reduction in LOS achieved by the program, 2) the per-day cost savings from decreased LOS, 3) the costs of implementing the program, and 4) the annual number of ICU admissions. The model was developed using Excel software (Microsoft Corporation, Redmond, WA), based on data from existing publications and available data from actual implementation of an early rehabilitation program in the Medical ICU (MICU) at Johns Hopkins Hospital. Variables for the Financial Model Length of Stay. Table 1 provides data on the average LOS prior to implementing an early rehabilitation program and the projected proportionate reductions in average LOS that were used in the example scenario and in the sensitivity analyses (i.e., bestcase and conservative-case scenarios) in the nancial model. The example scenario is based on actual data from the MICU at Johns Hopkins Hospital. The best-case and conservativecase sensitivity analyses were designed to present reasonable optimistic and pessimistic nancial outcomes, respectively. For the conservative scenario, ICU and oor LOS were drawn from a study of 271 ICUs in 188U.S. nonfederal hospitals (15), with oor LOS estimated by subtracting the average ICU LOS from the average total hospital LOS. For the best-case scenario, ICU and oor LOS were assumed to be modestly higher than the actual average LOS in the example scenario (at the Johns Hopkins MICU) prior to initiation of the early ICU rehabilitation program. Preintervention average LOS and LOS percent reductions due to an early rehabilitation program in the conservativeand best-case scenarios (Table 1) are based on published data (6, 810, 16). These data demonstrate ICU LOS reductions of approximately 20%38% (6, 810, 16). Reductions in oor LOS have been 10%25% (6, 9), with one randomized controlled trial suggesting a decrease in ICU average LOS without a decrease in average total hospital LOS (8). For the example scenario, based on Johns Hopkins MICU data, the LOS reductions are an average of the actual LOS reductions achieved when comparing 3-month time periods (April to June) in 2006 and 2008, in which there was no rehabilitation program, vs. the same 3-month time periods in 2007 and 2009, in which there TABLe 1.Assumptions

was a rehabilitation program (initially a quality improvement project in 2007 and then a hospital-funded rehabilitation program from 2009 onwards), as illustrated in Figure 1 (9, 17). This specic April-June 3-month time period was compared over the 4-yr period in performing the LOS analysis since it prevented any seasonal effect from affecting the LOS comparisons and permitted the creation of two time periods with a rehabilitation program and two time periods without a rehabilitation program. Cost Savings From Decreased LOS. In estimating the cost savings attributable to LOS reductions, it was recognized that certain costs are xed over the short or medium term, such as salaries and benets for full-time staff, facilities, and equipment (18). Hence, as explained elsewhere (19), the estimate of cost savings from a reduced LOS that is most conservative (i.e., generates a lower value for estimated cost savings) uses directvariable costs. Direct-variable costs are used to estimate cost savings generated from not supplying medical services, including patient consumable costs, such as materials and services from blood bank, laboratory, pharmacy, radiology, and respiratory care (19). Since actual direct-variable cost estimates for ICU and oor days were not available for the example scenario, the values used in this evaluation are based on a prior detailed nancial analysis (19), which demonstrated decreasing directvariable costs over the rst 5 days of an ICU stay (Table 2), but relatively similar costs over the duration of a oor stay (oor cost in Table 2 is an average of the published values available for the rst 3 days (19)). These costs are substantially lower than other published costs (20), leading to a conservative bias for the projections calculated in the nancial analysis. Furthermore, we excluded the nancial benet of additional net revenue that may be generated from new admissions occurring due to increased capacity in the ICU and oor (due to decreased LOS), further resulting in a conservative bias that may understate the actual net cost savings. Annual Number of ICU Admissions. The nancial analysis accounted for the annual number of ICU admissions at a given hospital, with 200, 600, and 2,000 admissions used in the sensitivity analyses in order to represent small, medium, and very large ICUs. The actual annual number of MICU admissions (approximately 900) at Johns Hopkins was used as the example scenario to represent a large ICU in the model.

for ICU and Hospital Length of Stay for Financial Model

Conservative-Case Scenarioa Example Scenariob
5.4 22% 10.3 19%

Best-Case Scenarioa
6.5 25% 11.5 25%

ICU average LOS (days) Reduction in average ICU LOS Floor average LOS (days) Reduction in average oor LOS

3 10% 5 10%

LOS = length of stay. a Conservative-case scenario based on published data (15) and best-case scenario based on data that are modestly higher than actual data from the example scenario that represents the Johns Hopkins Medical ICU prior to initiation of the rehabilitation program. b The average ICU and oor LOS for the example scenario uses actual data from the Johns Hopkins Medical ICU for the 12-mo period from April 2006 to March 2007. The reduction in ICU and oor LOS are described in footnote b of Figure 1.

718

www.ccmjournal.org

March 2013 Volume 41 Number 3

Feature Article

Figure 1. Average length of stay with and without a rehabilitation program at Johns Hopkins Medical ICU (MICU)a,b. Rehab = Rehabilitation; QI = Quality Improvement; Prog. = Program. a Each data point represents the same 3-mo period over 6 consecutive years in order to allow for seasonal comparability between years with and without early rehabilitation interventions (i.e., no rehabilitation in 2006 and 2008 vs. the 2007 rehabilitation quality improvement project and the hospital-funded MICU rehabilitation program from 2009 onwards [9]). b The percentage decrease in length of stay (LOS) comparing before vs. after implementation of early rehabilitation was calculated by comparing the average LOS of 2006 and 2008 (no rehab periods) vs. the average LOS of 2007 and 2009 (intervention periods), using data in Panels A and B, as follows: 1) MICU LOS: {[(6.0 + 6.5) / 2] [(4.7 + 5.0) / 2]} / [(6.0 + 6.5) / 2] = 22%, 2) Hospital LOS: {[(16.5 + 19.5) / 2] [(12.9 + 15.9) / 2]} / [(16.5 + 19.5) / 2] = 20%, 3) Floor LOS (estimated as Hospital LOS MICU LOS):

{[(16.5 6.0) + (19.5 6.5)]/ 2} {[(12.9 4.7) + (15.9 5.0)]/ 2} = 19% [(16.5 6.0) + (19.5 6.5)]/ 2
The number of MICU admissions during these four 3-mo periods from 2006 to 2009 were 210, 237, 176, and 251, respectively, with the average percent increase in MICU admissions comparing before vs. after implementation of early rehabilitation calculated, similarly to above, as {[(237 + 251) / 2] [(210 + 176) / 2]} / [(210 + 176) / 2] = 27%.

Critical Care Medicine

www.ccmjournal.org

719

Lord et al

TABLe 2. Daily

Direct-Variable Costs, by Day, for ICU and Floor

Daily ICU Direct-Variable Costs by Day Conservative-Case Scenario Example Scenario 100% of Costa
$3,768 $1,057 $839 $834 $690

Best-Case Scenario 120% of Costa

$5,275 $1,480 $1,175 $1,168 $966

Day of Stay
Day 1 Day 2 Day 3 Day 4 Day 5+

80% of Costa
$2,261 $634 $503 $500 $414

Daily Floor Direct-Variable Costs 80% of Costa

All days
a

100% of Costa
$249

120% of Costa
$299

$199

Cost is based on published data (19).

Costs. The nancial model accounted for variability in the cost of implementing an ICU rehabilitation program based on the annual number of ICU admissions, as described above. There are three main categories of costs associated with an ICU rehabilitation program: 1) personnel, 2) training, and 3) equipment (Table 3). Personnel: Physical therapists (PTs) and/or occupational therapists (OTs), rehabilitation aides/technicians, a program coordinator, and a physician leader are personnel commonly involved in an ICU rehabilitation program. Their typical roles are briey described herein with further details provided elsewhere (14, 21). Use of these personnel may vary between programs (e.g., some programs may not include funding for a specic category of personnel), but their costs were included TABLe 3. Estimated

in the nancial model in order to provide a conservative estimate of costs for a comprehensive program (22). The number of full-time equivalent staff is based on actual experience with the early rehabilitation program at the Johns Hopkins MICU and data from other existing publications (6, 8, 12) with proportionate upward and downward adjustments made for varying ICU sizes. The types and duration of ICU rehabilitation therapy interventions vary with patients physiological stability and functional status, with interventions ranging from passive range of motion to assisted ambulation and treatment durations ranging from 1060 mins, provided once or twice daily (6, 8, 9, 2325). In addition to PT/OT staff that directly provides early rehabilitation therapy, a rehabilitation aide/

Costs of Rehabilitation Program by Number of ICU Admissions

200 600 900 2000

Salary and fringe benets Physical therapist/occupational therapist (PT/OT)a Rehabilitation aide/technician Coordinator Training Equipmente Total cost
a b

$34,134 $0 $0 $0 $0 $0 $34,134

$155,153 $12,904 $24,090 $10,230 $2,000 $20,000 $224,378

$206,871 $39,104 $73,000 $15,500 $4,000 $20,000 $358,475

$413,743 $78,208 $73,000 $31,000 $10,000 $40,000 $645,951

c d

ICU physician leader

Assuming 0.33, 1.5, 2.0, and 4.0 full-time equivalent PTs/OTs for 200, 600, 900, and 2,000 annual admissions, respectively. Assuming 0, 0.33, 1.0, and 2.0 full-time equivalent technicians. c Assuming 0, 0.33, 1.0, and 1.0 full-time equivalent coordinator. d Part-time effort from ICU physician for leadership and medical directorship of program. e Cost estimate generated assuming purchase of one portable ventilator ($15,000) and two wheelchairs for patient seating out of bed (total $5,000) for 600 and 900 annual ICU admissions, and two portable ventilators and four wheelchairs at 2,000 admissions. The actual equipment required in a specic ICU will vary from this based on available equipment. These equipment costs represent a maximally conservative assumption in the nancial analysis since the entire cost is attributed to a single year without any depreciation of the cost over the useful life of the equipment.

720

www.ccmjournal.org

March 2013 Volume 41 Number 3

Feature Article

technician can play a valuable role on an early ICU rehabilitation team. Often, more than one person may be required to manage critically ill patients and their medical equipment during rehabilitation sessions. In such cases, the rehabilitation aide/technician can assist and increase the efciency of more costly PT/OT staff during therapy (14). Furthermore, a technician may play a role in providing other interventions, such as passive range of motion in comatose or sedated patients, which have frequently been included in early rehabilitation protocols (6, 8). A program coordinator can play many different roles in an early ICU rehabilitation team. Responsibilities may include screening ICU patients to identify, on a timely basis, appropriate patients for initiation of early rehabilitation therapy using predened criteria (6, 8, 9, 12, 25, 26), so that the PT/OT personnel may focus predominantly on provision of rehabilitation therapy. The coordinator can also assist with ongoing program evaluation, important aspects of ICU culture change, and the quality improvement efforts that are essential to a successful program (10, 17, 27). In addition, part-time effort from an ICU physician leader can play an important role in providing medical input regarding rehabilitation of complex critically ill patients and education to ICU providers regarding the importance of early mobilization activities, which is necessary for culture change. Training: Additional training, including conferences, courses, and site visits to existing early rehabilitation programs, are important to assist clinicians in gaining the skills and insights necessary for creating and maintaining an early rehabilitation program. Equipment: There is a wide range of standard and customized rehabilitation equipment that can assist with ICU patient mobilization activities, as described elsewhere (14). The equipment costs included in the nancial model allows for the purchase of such equipment, if the ICU is not pre-equipped with it. Since the authors cannot account for the availability of existing equipment in any given ICU, the model includes some reasonable example costs, as described in Table 3, based on experiences with the program in the Johns Hopkins MICU. Calculation of Net Financial Savings and Sensitivity Analyses Table 4 illustrates the nancial model that uses the above variables to calculate the annual net savings (or costs) associated with an ICU rehabilitation program. To ensure that the models results were robust to varying key assumptions, a sensitivity analysis was designed and conducted. This sensitivity analysis simultaneously varied the following factors: 1) the direct-variable costs of an ICU and oor day (80%, 100%, and 120% of published values (19), as per Table 2), 2) the estimated costs of a rehabilitation program based on the annual number of ICU admissions (Table 3), and 3) the reduction in ICU and oor LOS using previously described best-case and conservative-case scenarios (Table 1). This Excel spreadsheet model, available upon request from the corresponding author, allows institutions to further tailor the nancial analysis beyond these sensitivity analyses, to replicate their specic situation and perform additional analyses. Critical Care Medicine

RESULTS
Using the nancial model, the example scenario based on actual LOS data from the Johns Hopkins MICU revealed a net cost savings of $817,836, due to $1,176,312 of savings attributable to projected reductions in direct-variable costs due to reduced LOS, partially offset by an investment of $358,475 to implement the ICU early rehabilitation program (Table 4). A sensitivity analysis of these ndings, varying the number of admissions, direct-variable costs per day, and LOS reductions, revealed net cost savings in 20 (83%) of the 24 different scenarios, with projections ranging from approximately $88,000 (net cost) to $3,763,000 (net savings) (Table 5). Under the best-case scenario assumptions for LOS, the savings were projected to consistently increase with the number of ICU admissions (range: $260,000$3,763,000; Table5). Even maximally conservative assumptions, which simultaneously combined the conservative-case scenarios for LOS and the lowest directvariable cost reduction estimates (i.e., using only 80% of published values (19)), resulted in relatively small net costs (range: $44,000$88,000; Table 5) except in the smallest ICUs. These smallest ICUs always had net savings, due to the low projected start-up and equipment costs of a very small program and the assumption that PTs and/or OTs could work on a part-time basis rather than being dedicated to an ICU, as frequently occur in larger sized programs (Table 3).

DISCUSSION
This nancial analysis projects that implementation of an ICU early rehabilitation program can result in substantial savings across a wide range of ICU sizes, based on cost savings from reductions in ICU and oor LOS that have been demonstrated in prior publications, including in our own experience at the Johns Hopkins Hospital MICU (611). Even when simultaneously combining all of our most conservative assumptions, the net cost of implementing an early ICU rehabilitation program remains modest, especially when considering the demonstrated improvements in patient outcomes consistently associated with early ICU rehabilitation (611). Without performing detailed nancial modeling, prior studies have suggested that early rehabilitation programs may be cost saving or cost-neutral via reduced LOS (6, 9, 10). Furthermore, early rehabilitation has demonstrated improvements in ICU patients muscle strength, physical function, and quality of life, and reductions in institutionalization, mortality, and hospital readmissions after discharge (611). These improvements in patient outcomes alone could justify the use of hospital resources, but this nancial analysis suggests that early rehabilitation may be a dominant strategy (28) because of both net cost savings and improved patient outcomes. Even with maximally conservative assumptions by simultaneously combining the conservative-case scenario for LOS and directvariable costs estimates of only 80% of published values (19), the net cost of early rehabilitation interventions remains modest given that decreased direct-variable costs offset most of
www.ccmjournal.org

721

Lord et al

TABLe 4. Financial Row

A B C D E F G H I J K L M N O P Q R S

Model for ICU Rehabilitation Program Using Example Dataa

Description Number
5.40 10.30 22% 19% 1.19 1.96 4.21 8.34 $7,464 $2,565 $10,029 $6,644 $2,077 $8,722 $1,307 900 $1,176,312 $358,475 $817,836 QR PO Table 3 AC BD AE BF [see Methodsc] [see Methodsc] I+J [see Methods] [see Methods] L+M K-N Table 1 Table 2 (using 100% ICU Costs) Table 2 (using 100% Floor Costs) Table 2 (using 100% ICU costs) Table 2 (using 100% oor costs)

Calculationsb

References
Table 1 Table 1 Table 1 Table 1

Average ICU length of stay (LOS) before intervention Average oor LOS before intervention % Reduction in ICU LOS % Reduction in oor LOS Average reduction in ICU LOS (in days) Average reduction in oor LOS (in days) Average ICU LOS after intervention Average oor LOS after intervention ICU direct-variable cost for patient with average LOS before intervention Floor direct-variable cost for patient with average LOS before intervention Total direct-variable cost for patient with average ICU and oor LOS before intervention ICU direct-variable cost for patient with average LOS after intervention Floor direct-variable cost for patient with average LOS after intervention Total direct-variable cost for patient with average ICU and oor LOS after intervention Total average cost savings per patient Annual number of ICU admissions Total cost savings across all admissions Costs for rehabilitation program Net savings associated with ICU rehabilitation program

LOS = length of stay. a The nal savings presented in this table are based on data derived from actual length of stay reductions achieved with the John Hopkins Hospital Medical ICUs early rehabilitation program. These nal savings exclude the effects of increased ICU and oor admissions from decreased LOS which could result in additional net revenue or the long-term cost savings from reduced personnel and nonpersonnel charges due to decreased patient-days of care. Furthermore, results of the model under-estimate potential cost savings due to allocating the entire purchase price of equipment to a single year without any depreciation of the cost over the useful life of the equipment. b The results of arithmetic calculations presented in this table may differ from actual calculations due to rounding. c Calculated as follows (using cost data from Table 2): the total ICU direct-variable cost for an ICU LOS of 5.4 days (Row A) is the sum of ICU costs for Day 1 ($3,768), Day 2 ($1,057), Day 3 ($839), Day 4 ($834), and 1.4 days of Day 5+ costs (1.4 $690), to provide a total ICU stay cost of $7,464, and the total oor direct-variable cost is the product of 10.3 days (Row B) and the per-oor day cost of $249, yielding a total oor stay cost of $2,565.

the increase in rehabilitation costs and given the context of improved patient outcomes. In several ways, we were strongly conservative in creating the nancial model to avoid overstating the potential net cost savings. First, many of the assumptions used in the model are intentionally conservative. For example, in our sensitivity analysis of net cost savings, the best-case scenario is much less optimistic than the conservative-case scenario is pessimistic. Indeed, the best-case scenario (Table 1) is very similar to the actual results achieved in the Johns Hopkins Hospital MICU. Second, potential net revenue from increased ICU capacity was not considered in the nancial analysis; admissions data at the Johns Hopkins MICU demonstrated a 27% average increase in the number
722
www.ccmjournal.org

of admissions in periods with early rehabilitation when compared to periods without early rehabilitation (Fig. 1). Similarly, this model also ignored any long-term cost savings that could occur, via reduced stafng and infrastructure needs, if available ICU beds were not occupied with new patients. Instead, this model only considered cost savings based on the direct-variable costs of consumable goods and services. Third, equipment costs were not depreciated over their useful life (29); their entire cost was allocated to 1 yr. Finally, no consideration was given to improved patient outcomes, which have been consistently demonstrated in published evaluations of ICU rehabilitation programs (611). Our study has potential limitations. First, we limited the scope of our analysis to only the nancial aspects of
March 2013 Volume 41 Number 3

Feature Article

TABLe 5. Sensitivity Analysis of Net Financial Savings (in $000s), by Number of Admissions, Direct-Variable Costs per Day, and Reduction in Length of Staya Number of Admissions 200
Direct-variable costs as % of published values (19) (Table 2) Best-case scenario for length of stay reduction (Table 1) Conservative-case scenario for length of stay reduction (Table 1)
a

600
120% 80%

900
100% 120% 80%

2000
100% 120%

80% 100% 120% 80% 100%

$260

$333

$407 $657

$878 $1,098 $964 $1,295 $1,626 $2,293 $3,028 $3,763

$26

$41

$56

-$44

$1

$46

-$88

-$20

$48

-$44

$106

$257

The net nancial savings (or costs, as represented by negative values) presented in this table do not include the effects of increased ICU and oor capacity from decreased length of stay which could result in additional net revenue from increased admissions or long-term cost savings from reduced personnel and nonpersonnel costs due to decreased patient-days of care.

implementing an ICU early physical rehabilitation program, without considering the clinical benets to patients that have previously been published and are cited herein. Second, in our estimated costs for ICU rehabilitation programs at hospitals of varying sizes, we assumed that the start-up costs and ongoing operational costs for ICU rehabilitation programs are relatively lower at smaller ICUs, due to rehabilitation staff not being dedicated to the ICU and limited investment in additional equipment as we have observed in other settings. Third, existing studies providing data for estimated LOS reductions are drawn primarily from mechanically ventilated patients in medical or mixed ICUs. Hence, the results may not be generalizable in specialty ICU settings where the average LOS is very short and little LOS reduction can be achieved (e.g., fast-track cardiac surgery patients) and may not reect the full range of heterogeneity in patients in an ICU with respect to acuity of illness, sedation status, and mortality. However, the LOS reductions presented using our medical ICU example were generated from analyses of all patients admitted to the MICU, including nonventilated patients, patients with variable acuity, and survivors and nonsurvivors, which helps support the generalizability of our results. We also assumed that all early rehabilitation interventions will reduce LOS, at least modestly, based on the experience of existing publications in this eld; however, ad hoc attempts to implement early rehabilitation as part of routine clinical practice may not decrease LOS or improve patient outcomes. To overcome this challenge, structured quality improvement models for introducing early ICU rehabilitation are available to help ensure success (9, 10, 17, 30). Furthermore, the net cost savings that we project for the example scenario do not represent actual net cost savings data. The savings projections were calculated based on actual LOS reductions, but used direct-variable cost data from a published source (19), as these data were not available for the Johns Hopkins MICU; however, the projected net cost savings are conservative and likely understate the actual savings that could be achieved. Critical Care Medicine

Another potential criticism is that the projected LOS savings may be due to changes in sedation practice, rather than early rehabilitation programs; however, in a randomized controlled trial of early PT and OT in which both the intervention and control groups had the same approach to sedation, a decrease in ICU LOS was demonstrated (8). Furthermore, in our actual experience in the Johns Hopkins Hospital MICU (Fig. 1), the decrease in ICU and hospital LOS during the fourth time period (the April-June 2009 intervention period) vs. the third time period (the April-June 2008 control period) occurred in the setting of sedation practices having already been changed (and maintained) 2 yr prior during the initial quality improvement project (April-June 2007), with the incremental change in the fourth period only being the reintroduction of additional rehabilitation stafng. An additional potential limitation is that our modeled cost projections may be applicable only to U.S. ICUs, because the per-day direct-variable cost data were derived from a U.S. ICU (19) and other countries have different hospital funding models. Finally, our experience, and much of the published literature, has focused on early rehabilitation in MICUs within academic medical centers. However, a meta-analysis of randomized trials has demonstrated that across a wide variety of acute care settings with different patient populations, additional physical therapy treatment signicantly reduces LOS and improves functional outcomes (31). Nonetheless, we encourage further study of early rehabilitation interventions in other settings.

CONCLUSIONS
A nancial analysis model, created using conservative assumptions based on actual experience and published data, projects that investment in ICU early rehabilitation programs can generate substantial net cost savings for hospitals while improving patient outcomes. Even under the most conservative assumptions, the projected net cost is modest relative to the substantial improvements previously demonstrated in patient outcomes.
www.ccmjournal.org

723

Lord et al

These analyses and the accompanying model should help address nancial barriers to implementation of early physical rehabilitation programs in ICUs.

REFERENCES

1. Desai SV, Law TJ, Needham DM: Long-term complications of critical care. Crit Care Med 2011; 39:371379 2. Herridge MS, Tansey CM, Matt A, et al: Functional disability 5 years after acute respiratory distress syndrome. N Engl J Med 2011; 364:12931304 3. Herridge MS, Cheung AM, Tansey CM, et al: One-year outcomes in survivors of the acute respiratory distress syndrome. N Engl J Med 2003; 348:683693 4. Barnato AE, Albert SM, Angus DC, et al: Disability among elderly survivors of mechanical ventilation. Am J Respir Crit Care Med 2011; 183:10371042 5. Bienvenu OJ, Colantuoni E, Mendez-Tellez PA, et al: Depressive symptoms and impaired physical function after acute lung injury: A 2-year longitudinal study. Am J Respir Crit Care Med 2012; 185:517524 6. Morris PE, Goad A, Thompson C, et al: Early intensive care unit mobility therapy in the treatment of acute respiratory failure. Crit Care Med 2008; 36:22382243 7. Morris PE, Grifn L, Berry M, et al: Receiving early mobility during an intensive care unit admission is a predictor of improved outcomes in acute respiratory failure. Am J Med Sci 2011; 341:373377 8. Schweickert WD, Pohlman MC, Pohlman AS, et al: Early physical and occupational therapy in mechanically ventilated, critically ill patients: A randomised controlled trial. Lancet 2009; 373:18741882 9. Needham DM, Korupolu R, Zanni JM, et al: Early physical medicine and rehabilitation for patients with acute respiratory failure: A quality improvement project. Arch Phys Med Rehabil 2010; 91:536542 10. Hopkins RO, Spuhler VJ, Thomsen GE: Transforming ICU culture to facilitate early mobility. Crit Care Clin 2007; 23:8196 11. Burtin C, Clerckx B, Robbeets C, et al: Early exercise in critically ill patients enhances short-term functional recovery. Crit Care Med 2009; 37:24992505 12. Bailey P, Thomsen GE, Spuhler VJ, et al: Early activity is feasible and safe in respiratory failure patients. Crit Care Med 2007; 35:139145 13. Pronovost PJ, Needham DM, Waters H, et al: Intensive care unit physician stafng: Financial modeling of the Leapfrog standard. Crit Care Med 2004; 32:12471253 14. Korupolu R, Sussai I, Needham DM: Series on early mobilisation of critically ill patients, Part three: Required resources. ICU Management 2010; 10:3637 15. Lilly CM, Zuckerman IH, Badawi O, et al: Benchmark data from more than 240,000 adults that reect the current practice of critical care in the United States. Chest 2011; 140:12321242

16. Malko M, Karadibak D, Yildirim Y: The effect of physiotherapy on ventilatory dependency and the length of stay in an intensive care unit. Int J Rehabil Res 2009; 32:8588 17. Needham DM, Korupolu R: Rehabilitation quality improvement in an intensive care unit setting: Implementation of a quality improvement model. Top Stroke Rehabil 2010; 17:271281 18. Wunsch H, Gershengorn H, Scales DC: Economics of ICU organization and management. Crit Care Clin 2012; 28:2537, v 19. Kahn JM, Rubenfeld GD, Rohrbach J, et al: Cost savings attributable to reductions in intensive care unit length of stay for mechanically ventilated patients. Med Care 2008; 46:12261233 20. Dasta JF, McLaughlin TP, Mody SH, et al: Daily cost of an intensive care unit day: The contribution of mechanical ventilation. Crit Care Med 2005; 33:12661271 21. Korupolu R, Gifford J, Needham DM: Early mobilization of critically ill patients: Reducing neuromuscular complications after intensive care. Contemporary Critical Care 2009; 6:112 22. Kho ME, Korupolu R, Needham DM: Early mobilisation of critically ill patients: Feasibility and benets. ICU Management 2009; 9: 3132 23. Gosselink R, Bott J, Johnson M, et al: Physiotherapy for adult patients with critical illness: Recommendations of the European Respiratory Society and European Society of Intensive Care Medicine Task Force on Physiotherapy for Critically Ill Patients. Intensive Care Med 2008; 34:11881199 24. Thomas AJ: Exercise intervention in the critical care unitWhat is the evidence? Physical Therapy Reviews 2009; 14:5059 25. Korupolu R, Chandolu S, Needham DM: Series on early mobilisation of critically ill patients: Part one: Screening and safety issues. ICU Management 2009; 9:2730 26. Stiller K, Phillips AC, Lambert P: The safety of mobilisation and its effects on haemodynamic and respiratory status of intensive care patients. Physiother Theory Pract 2004; 20:175185 27. Bailey PP, Miller RR 3rd, Clemmer TP: Culture of early mobility in mechanically ventilated patients. Crit Care Med 2009; 37(10 Suppl):S429S435 28. Tilleul P, Aissou M, Bocquet F, et al: Cost-effectiveness analysis comparing epidural, patient-controlled intravenous morphine, and continuous wound inltration for postoperative pain management after open abdominal surgery. Br J Anaesth 2012; 108:9981005 29 Libby PA, Libby R, Short DG: Financial Accounting: A Global Perspective. Fifth Edition. New York, McGraw-Hill, 2007 30. Pronovost PJ, Berenholtz SM, Needham DM: Translating evidence into practice: A model for large scale knowledge translation. BMJ 2008; 337:a1714 31. Peiris CL, Taylor NF, Shields N: Extra physical therapy reduces patient length of stay and improves functional outcomes and quality of life in people with acute or subacute conditions: A systematic review. Arch Phys Med Rehabil 2011; 92:14901500

724

www.ccmjournal.org

March 2013 Volume 41 Number 3