OLDER PEOPLE

A randomized controlled trial of a community nurse-supported hospital discharge programme in older patients with chronic heart failure
Timothy Kwok
MD, FRCP

Department of Medicine & Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China

Jenny Lee

MSc, MRCP

Department of Medicine & Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China

Jean Woo

MD, MA, FRCP

Department of Medicine & Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China

Diana TF Lee

RN, PhD

The Nethersole School of Nursing, The Chinese University of Hong Kong, Hong Kong, China

Sian Grifth

MA, FRCP

School of Public Health, The Chinese University of Hong Kong, Hong Kong, China

Submitted for publication: 20 June 2006 Accepted for publication: 22 December 2006

Correspondence: Professor Timothy Kwok Department of Medicine & Therapeutics Prince of Wales Hospital Hong Kong China Telephone: (852)-26076900 E-mail: tkwok@cuhk.edu.hk

K W O K T , L E E J , W O O J , L E E T F D & G R I F F I T H S ( 2 0 0 8 ) Journal of Clinical Nursing 17, 109117 A randomized controlled trial of a community nurse-supported hospital discharge programme in older patients with chronic heart failure Aims and objectives. To evaluate the effectiveness and cost-effectiveness of a community nurse-supported hospital discharge programme in preventing hospital re-admissions, improving functional status and handicap of older patients with chronic heart failure. Design. Randomized controlled trial; 105 hospitalized patients aged 60 years or over with chronic heart failure and history of hospital admission(s) in previous year were randomly assigned into intervention group (n 49) and control group (n 56) for six months. Intervention group subjects received community nurse visits before discharge, within seven days of discharge, weekly for four weeks, then monthly. Community nurse liaised closely with a designated specialist in hospital and were accessible to subjects during normal working hours. Control and intervention group subjects were followed up in the same specialist medical clinics. Primary outcome was the rate of unplanned re-admission at six months. Secondary outcomes were number of unplanned re-admissions, six-minute walking distance, London Handicap Scale and public health care and personal care costs. Results. At sixth months, the re-admission rates were not signicantly different (46 vs. 57% in control subjects, p 0233, Chi-square test). But the median number of re-admissions tended to lower in the intervention group (0 vs. 1 in control group,

2008 The Authors. Journal compilation 2008 Blackwell Publishing Ltd doi: 10.1111/j.1365-2702.2007.01978.x

109

T Kwok et al.

p 0057, Mann Whitney test). Intervention group subjects had less handicap in independence (median change 0 vs. 05 in control subjects, p 0002, Mann Whitney test), but there was no difference in six-minute walking distance. There was no signicant group difference in median total public health care and personal care costs. Conclusion. Community nurse-supported post-discharge programme was effective in preserving independence and was probably effective in reducing the number of unplanned re-admissions. The cost benets to public health care were not signicant. Relevance to clinical practice. Older chronic heart failure patients are likely to benet from post-discharge community nurse intervention programmes. More comprehensive health economic evaluation needs to be undertaken. Key words: cogestive, community health nursing, cost benet analysis, heart failure, older people

Introduction
Chronic heart failure (CHF) is a major health problem of older people and is associated with recurrent hospital admissions (Kwok et al. 1999). To account for this, several factors have been identied. Lack of social support and psychological maladjustment could contribute to symptoms and health seeking behaviour. Under use of evidence-based care for CHF, poor drug and dietary adherence, poor access to medical care are additional potentially avoidable factors (Williams & Fitton 1988, Ahmed 2003). Post-hospital discharge home visits by nurses, with or without medical or multidisciplinary support (Rich et al. 1995, Stewart et al. 1999, Blue et al. 2001, McAlister et al. 2004), have been shown to be effective in preventing readmission of high-risk patients with CHF. These programmes have also been shown to be cost-effective (Williams & Fitton 1988, Naylor et al. 1994, Phillips et al. 2004). However, these studies were performed in developed countries where older people enjoyed comprehensive coverage of primary and secondary health care either by insurance or by public funding. The publicly funded hospitals in Hong Kong are well equipped and their charges are nominal. On the contrary, primary health care is primarily private and not easily affordable to many older people. There is, therefore, a big incentive for the chronic sick to rely on the Accident and Emergency Departments (A&E) of public hospitals for the relief of their symptoms. The communication between hospital doctors and general practitioners is also poor. For example, on hospital discharge, patients are only given a summary of medical diagnoses and medication and the summaries are not directly sent to the general practitioners.
110

It is, therefore, not surprising that older CHF patients in Hong Kong have very high re-admission rates (at four weeks as high as 15%) (Kwok et al. 1999). In a local prospective study of hospital-discharged older medical patients, major problems in communication about medications and followup arrangement, lack of community support and increase in functional disabilities were identied (Woo & Cheung 1993). In view of these problems, we envisaged that the postdischarge community nursing programme for older CHF patients had to be more intensive in Hong Kong. We, therefore, performed a randomized controlled trial to examine its effectiveness and cost-effectiveness. The hypothesis was that an intensive post-discharge community nursing programme will reduce the chance of re-admission by improving functional status and reducing handicap of older CHF patients.

Methods
Subjects
Hospital patients with a principal diagnosis of CHF were recruited from the medical wards in Prince of Wales Hospital (PWH), a major teaching hospital in Hong Kong. A minority of subjects were recruited in another acute district general hospital, Alice Ho Miu Ling Nethersole Hospital (AHNH) in the same health region. The two hospitals were publicly funded and provided comprehensive specialist inpatient and outpatient services for a regional population of one million. The mean length of stay in the medical departments in these hospitals was ve days. Some frailer CHF patients, who required a longer period of convalescent care, were discharged to the two convalescent hospitals in the region

2008 The Authors. Journal compilation 2008 Blackwell Publishing Ltd

Older people

Post-discharge program in heart failure

Shatin and Taipo Hospitals. The inclusion criteria included age older than 60, residing within the region and had at least one hospital admission for CHF in the 12 months prior to the index admission. The exclusion criteria included having communication problems but without caregivers, residing in a nursing home and terminal diseases with a life expectancy of less than six months.

randomization grouping of the subjects. All hospital admissions, including attendance to the A&E, throughout Hong Kong were traced by an electronic database maintained by the Hospital Authority which operated all publicly funded hospitals in Hong Kong. Intervention group The subjects were visited by a designated community nurse (CN) before they were discharged from the hospital. The objectives were to provide health counselling, such as drug compliance, dietary advice and to encourage subjects to contact CN via a telephone hotline during ofce hours when they developed symptoms. The CN carried a pager and a mobile phone. The trained clerk, who answered the hotline, relayed the message from the subjects to the CN via the pager. The subjects were then visited by the CN at home within seven days of discharge. During the home visits, the CN checked vital signs and signs for poor control of CHF ankle swelling, dyspnoea and basal crepitation on auscultation. Medications were checked and compliance encouraged. Avoidance of salty and high fat foods and regular physical exercise were promoted. Home care and day care services were arranged if social support was found to be insufcient. The CN thereafter performed home visits at weekly intervals for another four weeks and monthly after that. The CN liaised closely with either a geriatrician or a cardiologist in their respective hospitals. After liaison, the CN could alter medication regime, arrange urgent hospital outpatient appointments and clinical admission. When subjects were re-admitted, the CN visited the patients in the hospital and provided background information to attending doctors. Subjects who refused further home visits were monitored by the CN by telephone. Control group The control subjects received usual medical and social care, except that they were followed up in the hospital outpatient clinics by the same group of designated geriatrician or cardiologist.

Procedure
Eligible subjects were identied and recruited by a research nurse (RN) on the day or the day before hospital discharge. After obtaining written consent from the subjects, the RN recorded demographic data. Functional status was assessed by six-minute walking test (Butland et al. 1982). Cognitive function, psychological state and handicap were assessed by Abbreviated Mental Test (AMT) (Chu 1999) General Health Questionnaire (GHQ) (Chi & Boey 1993) and London Handicap Scale (LHS) (Lo et al. 2001), respectively. General Health Questionnaire is a screening test aimed at detecting psychological problems in people living in the community. It has been translated into Chinese and validated among older Chinese in Hong Kong (Chi et al. 1995). The score ranges from 030, with scores of six or above indicating psychological problems. The LHS was developed in the United Kingdom and had been translated into Chinese and validated in Hong Kong (Lo et al. 2001). It consisted of one question for each of the six domains of handicap (mobility, independence, occupation, social, orientation and economic). For each domain, there were six progressive levels of handicap. Higher score indicates higher level of handicap. In subjects with AMT score less than 6/10, family caregivers provided information for the handicap scale. The ward nurses then phoned a second research assistant who assigned trial grouping according to a random number table. The group assignment was made known to the patients. All subjects were followed up by designated geriatricians or cardiologists in their respective hospital medical clinics. The interval of clinic appointment ranged from 612 weeks on average. When the subjects were re-admitted, they were assessed by either a geriatrician or a cardiologist for reasons of re-admission. The primary and secondary causes of re-admission were categorized as follows: exacerbation of existing disorder, new but related event, new unrelated event, elective, admission with no deterioration, social/psychological problems, drug-related problems and dietary non-compliance. After six months of trial, subjects had their functional and psychosocial status re-assessed, as in the baseline, at the follow-up clinics. The RN was not aware of the

Ethics approval
The study was approved by the Research Ethics Committee of the Chinese University of Hong Kong.

Sample size calculation

According to the results from a pilot study of 49 older CHF patients discharged from PWH, the re-admission rate at six
111

2008 The Authors. Journal compilation 2008 Blackwell Publishing Ltd

T Kwok et al.

months was 69%, a sample size of 50 per group was estimated to have 80% chance of detecting a 40% relative reduction in re-admission rate at a condence interval of 095.

Statistical analysis
The primary outcome was the percentage of subjects who ever had unplanned hospital re-admissions within six calendar months of discharge. The secondary outcomes included the following: number of unplanned hospital re-admissions, changes in six-minute walking test and LHS domain scores. Comparison of proportion of subjects with re-admissions was made by Chi-square test. As the secondary outcome variables were not normally distributed, Mann Whitney U test was used for group comparison. Because LHS has six domains, p-value has to be <0008 before statistical signicance is assumed.

visits to general practitioners either in government outpatient clinic and private practice. Costs of travelling and escort were included. The costs of hospital care were calculated by the standard charges. The charges for one day of hospital stay, A&E attendance, one specialist outpatient attendance, general outpatient attendance and one CN visit were HK$100, 0, 50, 50, 55, respectively. 2 Social care social services, including old age home care, which were paid for by the subjects own resources, were recorded by CN throughout the follow-up period. In the control subjects, this was enquired about at the end of six months follow-up. In subjects who received comprehensive social security allowance (CSSA) and who were current or retired civil servants, all public services were free.

Results
Between September 1999 and February 2001, 105 CHF subjects were randomized into intervention (n 49) and control (n 56) groups. Their baseline characteristics, cardiac function status, co-morbidities and use of heart failure medications were compared in Table 1. The intervention group subjects were more likely to be recipients of CSSA and had greater economical handicap. The ow diagram of the randomized subjects and their major outcomes are shown in Fig. 1. One intervention and two control group subjects dropped out because of moving out of Hong Kong and the development of symptomatic cancer. Four intervention and eight control subjects died. The intervention group received an average of 88 (SD 34) visits and 53 (SD 27) telephone calls during the study period. The average total nursing hour per subject was 150 (SD 55). The six-month re-admission rates of intervention and control group subjects were 46 and 57%, respectively. The difference was not signicant (p 0233, Chi-square test). The median number of unplanned re-admissions tended to be lower in the intervention group [0 (quartile range 0, 1) vs. 1 (0, 2) in control group, p 0057, Mann Whitney test]. Out of 126 unplanned re-admission episodes to PWH in CHF subjects, 61 episodes (48%) were assessed by the geriatric team. The primary reasons of re-admission were as follows: exacerbation of CHF (n 32, 52%); new but related events (n 8, 13%); new unrelated event (n 11, 18%); psychosocial problems (n 4, 7%); drug-related problems (n 4, 7%) and dietary non-adherence (n 2, 3%). There was no signicant group difference in the primary causes of re-admission.

Cost analysis
The cost analysis was based on intention-to-treat for the perspectives of the public health care system and that of the individual patients. Costs to public health care system 1 Community nursing The numbers of CN visits, including predischarge hospital visits and telephone calls for each subject were recorded. Time spent on each item during the study period of six months was individually recorded and the average time spent on each item was calculated. Travelling time was calculated to be 10 minutes per visit according to records from CN. The costs of these items were then estimated by multiplying the hourly salary of CN by the total number of hours spent on each subject. 2 Specialist consultation and emergency care attendances All medical consultations in medical/geriatric outpatient clinics, geriatric day hospitals, hospital wards and emergency care were counted. 3 Number of days in different types of hospital wards was counted. Hospital beds in Hong Kong were charged solely according to the type of bed (acute vs. rehabilitation) and duration of stay, disregarding costs for medications, investigations and therapy. 4 Government outpatient clinics all subjects were given a small record book to note down the dates and the locations of general outpatient visits if any. Cost to patients 1 Medical care all subjects or close family members were given a small diary to note down the date and the costs of
112

2008 The Authors. Journal compilation 2008 Blackwell Publishing Ltd

Older people Table 1 Baseline characteristics of intervention (community nursing) and control group subjects*

Post-discharge program in heart failure

Intervention (n 49) Male Age (years) Live alone CSSA Six-minute walking test (m) Abbreviated Mental Test (max. 10) General Health Questionnaire (max. 30) London Handicap Scale (max. 6) Mobility Independence Occupation Social Orientation Economic Left ventricular EF <40% Ischemic heart disease Myocardial infarction Obstructive airway disease Diabetes mellitus Atrial brillation Hypertension Medications Diuretic ACE inhibitor Digoxin Beta-blocker 22 (45%) 795 66 15 (31%) 23 (47%) 1207 620 86 14 56 37 3 (2, 3) 3 (2, 3) 3 (3, 4) 2 (2, 2) 2 (2, 2) 3 (3, 4) 9 (18%) 23 (48%) 9 (18%) 4 (8%) 14 (29%) 14 (29%) 26 (54%) 40 (82%) 29 (59%) 9 (18%) 9 (18%)

Control (n 56) 25 (45%) 768 70 10 (18%) 14 (25%) 1185 625 87 13 61 33 3 (3, 3) 3 (2, 3) 3 (3, 4) 2 (2, 2) 2 (2, 2) 3 (2, 3) 15(30%) 26 (46%) 15 (27%) 7 (13%) 21 (38%) 17 (30%) 21 (38%) 55 31 10 14 (98%) (55%) (18%) (25%)

*Continuous data presented as mean standard deviation or median (quartile range); categorical data as number (%). Two control and one intervention group subjects had missing data because of poor mobility. Six intervention and six control group subjects did not have echocardiogram. CSSA, comprehensive social security assistance; EF, ejection fraction; ACE, angiotensin converting enzyme.

Table 2 Comparison of changes in functional and psychosocial status in intervention (community nursing) and control group subjects at six-month follow-up* Intervention (n 44) Six-minute walking test (m) London Handicap Scale Mobility Independence Occupation Social Orientation Economic 44 (15, 84) 0 0 0 0 0 0 (1, 0) (1, 1) (1, 0) (0, 1) (0, 0) (0, 0) Control (n 46) 25 (22, 69) 0 (0, 0) 05 (0, 1) 0 (03, 1) 0 (0, 1) 0 (0, 0) 0 (1, 0)

*Change follow-up baseline value, expressed in mean (standard deviation). Two intervention and three control group subjects had missing data because of poor mobility. Signicant difference after adjustment for multiple comparisons, p < 0005, Mann Whitney.

At the six-month follow-up, 44 in the intervention group and 46 in the control group were available for re-assessment for the six-minute walking test and LHS scores (Table 2). When compared with the control subjects, the intervention group subjects became signicantly less limited in independence (median change in LHS independence domain score 0 vs. 05, p < 0005, Mann Whitney test). The change in functional status of the subjects, as reected by the sixminute walking test, was not different between the groups. The group comparison of costs to public health care system was shown in Table 3. One intervention group subject and two control group subjects had missing records of outpatient attendance and were therefore excluded. The median community nursing costs in the intervention group was HK$2 391 per subject. The median total public health costs as a result of hospital stay and emergency care attendances were signicantly lower in intervention group than in control group (HK$5 229 vs. HK$20 916, p 0048). However, the total public health care costs were not signicantly different.
113

2008 The Authors. Journal compilation 2008 Blackwell Publishing Ltd

T Kwok et al.

Randomized (n = 105)

Intervention (n = 49)

Control (n = 56)

Dropout (one moved )

Dropout (one moved, one cancer) Died (n = 8) Six month follow-up Completed (n = 44) Completed (n = 46)

Died (n = 4)

Not readmitted (n = 25)

Readmitted (n = 19)

Readmitted (n = 24)

Not readmitted (n = 22)

Figure 1 Flow diagram of major outcomes of randomized subjects.

Cost item Hospital bed and emergency care Outpatient clinic Community nursing Total

Intervention (n 48) 5 1 2 10 229 365 391 186 (0, 33 384) (910, 1 544) (1 600, 3 050) (3 785, 37 962)

Control (n 54) 20 916 (534, 72 312) 1 365 (533, 1 365) 0 (0, 0) 21 599 (1 978, 73 449)

Table 3 Comparison of total public health costs per person in six months between intervention (community nursing) and control groups*

*HK$ presented as median (interquartile range).

The medical and social costs paid by the subjects were shown in Table 4. One intervention and two control group subjects who attended the six month follow-up had missing personal cost data. There was no signicant group difference in the total health and social care costs to the patients. However, only 17 out of 49 intervention group subjects (34%) paid for CN visits: 24 by CSSA, one by civil service pension scheme, eight by research grant because of refusal to pay. To exclude the confounding effect of waivers, the personal cost analysis was repeated after excluding those eligible for CSSA or civil service pension in both groups and on the assumption that all those subjects who had to pay for CN paid by themselves. The median cost of CN visits per subject was HK$385 (quartile range 330, 578). The median total personal costs (medical and social included) of the intervention (n 21) and control (n 32) group subjects
114

Table 4 Medical and social care costs to subjects in six months* Intervention (n 43) 0 100 0 0 0 0 513 (0, 132) (38, 150) (0, 136) (0, 330) (0, 360) (0, 10) (213, 2 407)

Cost item Outpatient clinic Travel to clinics/hospital Hospital stay Community nurse Private doctor Social services Total

Control (n 44) 132 75 0 0 0 0 775 (0, 132) (33, 144) (0, 527) (0, 0) (0, 608) (0, 0) (184, 2 107)

*HK$ presented as median (25th, 75th percentile); 12 control and 22 intervention group subjects had all charges waived because of comprehensive social security assistance or retired civil servant status; seven intervention group subjects had community nurse visits paid by research fund. Home help, meals on wheel, day care centre, escort for medical follow-up.

2008 The Authors. Journal compilation 2008 Blackwell Publishing Ltd

Older people

Post-discharge program in heart failure

were HK$1 457 (589, 4 011) and 922 (193, 2 269), respectively. The difference was not signicant (P 0118, Mann Whitney U test).

Discussion
Our results demonstrated that an intensive post-discharge community nursing intervention programme for older CHF patients could not reduce the chance of re-admissions in six months, but was probably effective in reducing the number of unplanned re-admissions. The latter is consistent with the results of other similar trials overseas. The low drop out rate in the intervention group reects an overall good acceptance of the intervention. The subjects were well matched except that the intervention group had more nancial hardship. This might have had a negative inuence on the rate of public health care utilization in the intervention group. As expected for older CHF subjects, only a third of the subjects had systolic heart failure. Consistent with clinical practice at the time of the study, the great majority of CHF patients were on diuretics and the use of angiotensin converting inhibitor (ACEI) and beta blockers was limited. It is noteworthy that the use of ACEI and beta blockers has been associated with less re-admission in CHF patients (Fowler et al. 2001, Abarca et al. 2004). The change of medication after recruitment was unfortunately not documented in this study. However, as both groups were managed by the same specialist outpatient clinics, signicant differences in medication use were unlikely. Although the intervention did not signicantly reduce the chance of admission within six months, it reduced the number of unplanned admissions. The lower mortality rates of the intervention group subjects (83 vs. 148% in control subjects) suggested that the CN intervention did not delay hospital admissions when they were needed. A meta-analysis showed that post-discharge CN interventions, when combined with specialist medical support, can reduce mortality of CHF patients (McAlister et al. 2004). Community nurse intervention did not improve functional status of the CHF patients. However, relative to the control group who showed deterioration in the independence domain of handicap, CN intervention helped to maintain the level of independence. The educational and advisory role of the CN had probably helped the patients and the family caregivers to better manage the disease and its associated disabilities. The CN might have been effective in preventing some admissions by enhancing the interface between hospital services and the medical needs of CHF patients and possibly by improving the self-management of the disease by the

patients or the family members. There was still room for improvement as a few intervention group subjects were still re-admitted for dietary non-adherence and drug-related problems. The CN intervention can potentially be more effective if they can have training in two important areas of selfmanagement of CHF, namely diet and exercise. There is good evidence that regular physical exercise is benecial to the exercise tolerance of CHF patients (Smart & Marwick 2004), even in advanced age and in the home setting (Corvera-Tindel et al. 2004). CNs are in a good position to motivate older patients to perform physical exercises that suit their lifestyle. However, they require training in the prescription of physical exercise and information about locally available exercise programmes. Salty diet is a common problem in older people, particularly in those with CHF. Simple advice to restrict salt is usually not adequate to change behaviour in the long term (Gonzalez et al. 2005). More training in dietetic approaches in gradual salt restriction is needed. The high rate of re-admissions for reasons which may or may not be related to CHF suggests that this group of patients requires close medical supervision. Unfortunately, most subjects in both groups relied primarily on specialist outpatient clinics, but, the frequency of specialist outpatient clinic is necessarily limited because of the workload in hospitals and patients may not welcome that. The cost data showed that the cost of transport to outpatient clinic comprised a signicant proportion of their health care costs. Moreover, many older people have trouble nding their way round the hospital clinics, therefore requiring family members to escort them. Primary health care clinics which are more locally accessible should play a major role in the follow-up of these highrisk individuals in the community. This is indeed the case in most Western countries, but our data conrmed that the use of primary health care was minimal (data not shown). This highlights the under use or inaccessibility of primary health care for at risk older people. As in public hospitals, the government general outpatient clinics (GOPD) had low charges which were waived in those on CSSA, but to be seen at the clinics, one had to queue for a consultation quota early in the morning. This is not feasible for an ill older person. In the last two years, the Hospital Authority has taken over the administration of the GOPD. This opens up the possibility of arranging follow-up in GOPD upon hospital discharge and the sharing of clinical data via the HA territory-wise computerized client management system (CMS). Private clinics are widely available in Hong Kong, but this group of patients might have found them unaffordable. For those who can afford it, the private doctors can potentially be
115

2008 The Authors. Journal compilation 2008 Blackwell Publishing Ltd

T Kwok et al.

more helpful if they can have access to clinical information about the patients. At present, all hospital patients are provided with a computer-generated discharge note showing the diagnoses and medication. The detailed discharge summary is deliberately withheld from the patients because of condentiality issues. Patients can request medical reports but, they are charged a signicant fee and the procedures normally take a few weeks. This administrative culture of non-disclosure hinders the free ow of clinical information which is critically important if doctors in the community are expected to participate in chronic disease management. Pilot programmes allowing selected private general practitioners to gain access to their regular patients clinical information in the CMS via the internet are under way. The CN programme was probably effective in reducing the costs of hospital stay and emergency care attendance, but when the costs of CN were considered, the net gain in public health care costs was not signicant. In contrast, most overseas post-discharge CN programmes were shown to be cost effective (McAlister et al. 2004). Admittedly, this trial was under-powered to detect a small reduction in health care costs. Nevertheless, this CN programme might have been more cost effective if less home visits were made. The number of home visits in this programme was greater than in most previous trials which emphasized predischarge face-to-face hospital visits and post-discharge phone followups more (Naylor et al. 1994, Krumholz et al. 2002). Apart from the need to have an early home visit for an initial assessment and formulation of management plan, the need for further visits will largely depend on the resources and physical condition of the patients. The greater use of telephone follow-up and empowerment of patients and their families in disease management may reduce the need for home visits, therefore enhancing the cost-effectiveness of the programme. The costs to the patients were not overall increased by CN intervention. However, a CN visit cost HK$55 in Hong Kong at the time (the charge was increased to HK$80 in 2003). After excluding those who were eligible for waived charges, the cost of CN visits was a signicant proportion of the total health care costs paid by the patients. Further cost analysis conrmed that the CN intervention tended to increase the overall cost of care to the patients. It was, therefore, not surprising that a signicant proportion of subjects who had to pay refused to do so. In Hong Kong, only those with net asset of less than HK$30 000 were eligible for comprehensive social security assistance which included waiving of all charges from public services. However, there are many at risk older people who are not eligible for CSSA, but have meagre monthly incomes. This seriously questions the
116

rationale of charging for community health services which can potentially reduce overall health care costs. The strength of this trial is that it was the rst randomized controlled trial of CN intervention in CHF patients in Hong Kong where the primary health care is not well organized. The outcomes in functional status and handicap and cost analysis were considered at the same time. The main limitation was the small sample size which only allowed the detection of a major effect of CN intervention. The subjects had multiple medical problems and high mortality. In the cost analysis, the hospital and clinic costs were all standardized. The indirect costs to patients were not considered. More comprehensive cost analysis is warranted.

Conclusion
We concluded that post-discharge visits by CNs in older chronic cardiac failure patients were not effective in reducing the chance of re-admission within a six month period. But, they were effective in preserving independence and were probably effective in reducing the number of unplanned readmissions. More comprehensive health economic evaluation needs to be undertaken.

Acknowledgements
We would like to express our gratitude to the participating community nursing teams, Ms Eliza Lau, Dr Chan Chi Kin, and Prof John Sanderson. The research was funded by the Health Services Research Committee/Health Care & Promotion Fund (HSRC/HCPF) of Hong Kong.

Statement of Competing Interests

None.

Contributions
Study design: KT, LTFD; data collection and analysis: KT, WJ; manuscript preparation: KT, LJ, GS, WJ.

References
Abarca J, Malone DC, Armstrong EP & Zachry WM III (2004) Angiotensin-converting enzyme inhibitor therapy in patients with heart failure enrolled in a managed care organization: effect on costs and probability of hospitalization. Pharmacotherapy 24, 351357. Ahmed A (2003) American College of Cardiology/American Heart Association Chronic Heart Failure Evaluation and Management guidelines: relevance to geriatric practice. Journal of the American Geriatrics Society 51, 123126.

2008 The Authors. Journal compilation 2008 Blackwell Publishing Ltd

Older people Blue L, Lang E, McMurray JJ, Davie AP, McDonagh TA, Murdoch DR, Petrie MC, Connolly E, Norrie J, Round CE, Ford I & Morrison CE. (2001) Randomized controlled trial of specialist nurse intervention in heart failure. British Medical Journal 323, 715718. Butland RJA, Pang J, Gross ER et al. (1982) Two, six and 12 minute walking tests in respiratory disease. British Medical Journal 284, 16071608. Chi I (1995) Mental health of the old-old in Hong Kong. Clinical Gerontologist 15, 3144. Chu LW, Pei CKW, Ho MH & Chan PT. (1995) Validation of the abbreviated mental test (Hong Kong version) in the elderly medical patient. Hong Kong Medical Journal 1, 207211. Corvera-Tindel T, Doering LV, Woo MA, Khan S & Dracup K (2004) Effects of a home walking exercise program on functional status and symptoms in heart failure. The American Heart Journal 147, 339346. Fowler MB, Vera-Llonch M, Oster G, Bristow MR, Cohn JN, Colucci WS, Gilbert EM, Lukas MA, Lacey MJ, Richner R, Young ST & Packer M (2001) Influence of carvedilol on hospitalizations in heart failure: incidence, resource utilization and costs. U.S. Carvedilol Heart Failure Study Group. Journal of the American College of Cardiology 37, 16921699. Gonzalez B, Lupon J, Herreros J, Urrutia A, Altimir S, Coll R, Prats M & Valle V (2005) Patients education by nurse: what we really do achieve? European Journal of Cardiovascular Nursing 4, 107 111. Krumholz HM, Amatruda J, Smith GL, Mattera JA, Roumanis SA, Radford MJ, Crombie P & Vaccarino V (2002) Randomized trial of an education and support intervention to prevent readmission of patients with heart failure. Journal of the American College of Cardiology 39, 8389. Kwok T, Lau E, Woo J, Luk JKH, Sham A, Wong E & Lee SH (1999) The incidence of and risk factors for hospital readmission in older

Post-discharge program in heart failure medical patients in Hong Kong a cohort study. Journal of the Royal College of Physicians of London 33, 153156. Lo R, Harwood R, Woo J, Yeung F & Ebrahim S (2001) Crosscultural validation of the London handicap scale in Hong Kong Chinese. Clinical Rehabilitation 15, 177185. McAlister FA, Stewart S, Ferrua S & McMurray JJV (2004) Multidisciplinary strategies for management of heart failure patients at high risk for admission: a systemic review of randomized trials. Journal of the American College of Cardiology 44, 810819. Naylor M, Brooten D, Jones R, Ravizzo-Mourey R, Mezey M & Pauly M (1994) Comprehensive discharge planning for the hospitalized elderly: a randomised clinical trial. Annals of Internal Medicine 120, 9991006. Phillips CO, Wright SM, Kern DE, Singa RM, Shepperd S & Rubin HR (2004) Comprehensive discharge planning with post discharge support for older patients with congestive heart failure: a meta-analysis. Journal of the American Medical Association 291, 13581367. Rich MW, Beckham V, Wittenberg C, Leven CL, Freedland KE & Carney RM (1995) A multidisciplinary intervention to prevent the readmission of elderly patients with congestive heart failure. The New England Journal of Medicine 333, 11901195. Smart N & Marwick TH (2004) Exercise training for patients with heart failure: a systematic review of factors that improve mortality and morbidity. The American Journal of Medicine 116, 693706. Stewart S, Marley JE & Horowitz JD (1999) Effects of a multidisciplinary, home-based intervention on unplanned readmissions and survival among patients with chronic congestive heart failure: a randomized controlled study. Lancet 354, 10771083. Williams EI & Fitton F (1988) Factors affecting early unplanned readmission of elderly patients to hospital. British Medical Journal 297, 784787. Woo J & Cheung A (1993) A survey of elderly people discharged from hospital. Journal of Hong Kong Medical Association 45, 291297.

2008 The Authors. Journal compilation 2008 Blackwell Publishing Ltd

117