Family Practice Vol. 16, No. 3, 289-293
© Oxford University Press 1999
Can GP input into discharge planning result in better outcomes for the frail aged: results from a randomized controlled trial
Division of General Practice, Central Sydney Area Health Service, 37 Booth Street, BALMAIN NSW 2041,
a Department of General Practice, University of Sydney, 37A Booth Street, BALMAIN NSW 2041,
b Group General Manager, Royal Prince Alfred Hospital and Concord Repatriation & General Hospital, Central Sydney Area Health Service, Executive Suite, Concord Hospital, CONCORD NSW 2139 and
c Clinical Director, General, Geriatric and Rehabilitation Medicine, Central Sydney Area Health Service, Department of Geriatric Medicine, Concord Hospital, CONCORD NSW 2139, Australia.
McInnes E, Mira M , Atkin N, Kennedy P and Cullen J. Can GP input into discharge planning result in better outcomes for the frail aged: results from a randomized controlled trial. Family Practice 1999; 16: 289–294.
Received 11 August 1998; Revised 17 December 1998; Accepted 28 January 1999.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Objective. We aimed to assess whether GP input into discharge planning for high-risk aged in-patients admitted under the care of a geriatrician results in improved patient outcomes.
Methods. We conducted a prospective randomized controlled trial in Sydney, Australia. The subjects were 364 patients aged 60 years and over. The main outcome measures included community service referral, accommodation changes, length of stay, readmission rate, length of time to first readmission and patient satisfaction with discharge arrangements.
Results. No significant differences were found with regard to length of stay, readmission rates or time to first readmission. Test-group subjects were significantly more likely to be recommended for community services at discharge and to report that hospital personnel had discussed their discharge plan with them. Significantly more of the test group reported that their return home was well prepared.
Conclusions. Although GP pre-discharge visits did not alter the likelihood of ‘hard outcomes such as risk of readmission’, the results suggest that quality of care is enhanced amongst patients receiving a pre-discharge visit and that GPs can perform a key role in planning post-discharge care with other services.
Keywords. Discharge planning, elderly readmission, GP.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
GPs provide ongoing care to patients (and their families) in the community and have detailed knowledge and understanding of their patients' past medical history, social circumstances and levels of function. GPs in Australia have, over the last 30 years, played a decreasing role in the care of hospitalized patients, and the information held by GPs about patients is rarely utilized by hospital staff when planning a patient's discharge back into the community. Yet, involvement of the GP in discharge planning may result in better continuity of care and services between hospitals and GPs1 and improved communication between GPs and hospitals can assist in identifying patients that would benefit from increased support following discharge.2
Communication between hospital staff and GPs is often cited as a weakness in relation to discharge planning, with several studies reporting that GPs experience dissatisfaction both with hospital–GP communication and liaison practices,1,3 and that hospitals do not capitalize on GPs' knowledge of patients' social milieux. Reports of feeling alienated from the hospital system are not uncommon,1 and the appropriateness of discharge plans formulated without GP input has been questioned.3 Yet there appears to be willingness on the part of GPs to be involved in the discharge planning process,4,5 and GPs feel that they are well placed to improve the continuity of care between hospitals and primary care.1
Options for GP involvement in discharge planning range from being a patient advocate, without direct involvement in patient care,6 to allowing GPs admitting rights and input into patient management for specific conditions.7
Of relevance for the study reported in this paper is that inappropriate and early discharge from health care facilities has been implicated as a factor in higher readmission rates for older people,8 with poor communication between health professionals identified as a major cause of problems encountered by older patients once they have returned home.9 Strategies tested for improving discharge planning of geriatric patients are the use of a comprehensive discharge planning protocol, a geriatric consultation team, a geriatric evaluation unit and intensive post-discharge follow-up involving either allied health professionals or primary care teams.8–11
However, while these studies have examined approaches designed to improve discharge planning and post-discharge outcomes for the elderly, and it has been suggested that GPs are an essential part of a multidisciplinary discharge planning team, we can find no studies that have evaluated the impact of formal GP input in discharge planning for elderly patients.
We report here the results of a randomized controlled trial which investigated the impact of hospital visits for the purpose of discharge planning by GPs on their frail, aged patients.
The following hypotheses were tested. An invitation to the GP to visit his/her patient in hospital prior to discharge:
- alters the patient's discharge plans;
- renders patients more likely to view positively their return home; or
- results in fewer unplanned hospital readmissions within 26 weeks of hospital discharge and with a longer time interval from hospital discharge to first readmission.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Subjects, setting, recruitment
Ethical approval for all components of this study was obtained from the Central Sydney Area Health Service Ethics Committee prior to commencement of the study.
Patients admitted to the geriatric units of a district hospital and a teaching hospital between October 1994 and June 1996 were eligible to participate in the study if the attending geriatrician determined that the patient fulfilled one or more of the following criteria: multiple community service user, refusal of community and/or health services, high carer stress, patient/carer choice of discharge option other than recommended by the hospital geriatric team, rapid or frequent readmissions, poor mental status, and dependence in self-care and/or ambulation.
Mental status was assessed by the Mini-mental status examination which was scored from 0 (extreme cognitive impairment) to 30 (no cognitive impairment), with a score of 0–17 indicating moderate/severe cognitive impairment, 18–23 probable mild impairment and 24–30 minimal/no cognitive impairment.12 Functional status was measured by the Barthel's Activities of Daily Living scale,13 which gives a possible score of 0–100, with a score of 60 being a pivotal point for distinguishing between dependence and assisted independence. The other criteria outlined above were identified from the geriatrician's and/or geriatric team's knowledge of the patient as well as the patient's history as recorded in the medical records on admission. Suitable patients would be identified either during ward rounds or weekly case conferences approximately 7–10 days before the patient's discharge.
Informed consent was then obtained from the patient or, in cases of cognitive impairment, from the patient's carer (the person identified as assuming primary responsibility for the patient's care after discharge).
Randomization procedure
Randomization of patients to receive either a GP visit in addition to standard discharge planning (test group) or standard discharge planning alone (control group) generally occurred 1–2 days after the patient was identified by the geriatrician and was carried out by an independent research officer using the sealed envelope method. A computer-generated sequence of randomization had previously been generated. Following allocation, it was not possible to conceal group status from hospital staff, patients or research officers. An unequal randomization (2 test:1 control) was employed because of concerns that the required numbers in the test group might not have been attainable due to GP reluctance to participate, as suggested by a pre-project survey.5
Intervention
GPs of patients randomized to the test group were invited by the geriatrician to make a pre-discharge visit. Visits usually took place approximately 1–5 days after contact with the geriatrician. A consultation sheet was issued requesting written information from the GP specific to the individual patient, for example the GP's recommendations for post-discharge community service provision. The GP was able to talk to medical and allied health staff if required, had access to the patient's medical notes and was able to see the patient.
Data collection and main outcome measures
Patients were interviewed by a research officer (a registered nurse) immediately prior to discharge and in the patient's place of residence 6 and 12 weeks after discharge. A telephone interview was conducted 26 weeks after discharge.
Data collected included demographic information, functional status measured using the Modified Barthel's Activities of Daily Living scale13 and cognitive function measured using the Mini-Mental Status Examination (MMSE).12
In order to examine the potential for GPs to alter discharge plans, information about number and type of community services and domicile (categorized as supported, that is, nursing home or hostel resident, and non-supported, that is residing at home with or without community services) prior to admission and following discharge was obtained from medical records.
Satisfaction with discharge planning was assessed by a previously piloted standardized questionnaire administered by the interviewer prior to discharge and at 6-week follow-up. This examined whether patients were involved in discussion of their discharge plan while in hospital, their views on how well prepared their return home was and, for the test group, their views on the usefulness of the GP visit.
Readmission rates were determined by patient report at the follow-ups and confirmed by a review of hospital records. Length of stay of initial hospitalization and number of days to first readmission were obtained from hospital records.
Analysis
Data were stored and analysed in SPSSPC version 5.0.14 The impact of the intervention on outcomes was analysed on an ‘intention to treat’ basis using analysis of variance and covariance, logistic regression, chi-square and descriptive statistics at the two-tailed level of significance. A P < 0.05 was chosen for significance. When comparing groups controlling for extraneous covariates, an analysis of variance was implemented through multiple regression which included the treatment variable as well as Barthel's score as a covariate and independent variables. Regression either assumed normality for continuous outcome measures or was based on the logistic model for binary outcome measures.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Sample characteristics
Four hundred and twenty-seven patients were eligible: of these, 15 were lost in inter-hospital transfers and 17 could not give informed consent as they were too sick or confused and did not have a carer available. Thirty-one patients did not wish to participate. Three hundred and sixty-four patients consented, resulting in a response rate of 85%. Two hundred and five patients were randomized to intervention group and 159 to the control group.
The frailty of this group of in-patients is reflected in their mean age (81 years) and in the low mean scores on the MMSE (22) and Barthel's scale (74). Fifty-six per cent of the sample were female and 12% were from a non-English speaking background. Most patients lived in non-supported accommodation (Table 1
). The groups appeared well matched as a result of randomization, although the study groups differed significantly on Barthel's score (Table 1). In view of this difference all subsequent analyses were adjusted, treating the Barthel's score as a continuous measure.
|
Of the 205 patients randomized to receive a GP visit, 106 patients (52%) were actually visited by their GP. A detailed analysis of factors associated with GPs visiting their patients when invited has been reported elsewhere.5
At the 6-week follow-up, 47 of the sample had died, 15 were lost and 32 declined to continue in the study, leaving 270 participants. Between the 6- and 12-week follow-ups, 19 patients died, 7 were lost and 4 declined to continue, leaving 240 participants. Between the 12- and 26-week follow-ups, a further 23 had died and 9 were lost, leaving 208 participants at the end of the study. There was no difference in attrition rates between the test and control groups.
Alteration of discharge plans
At admission, there was no significant difference between the study groups in likelihood of receiving support services (OR = 1.43; 95% CI = 0.92–2.21; P = 0.10). However, significantly more of the test group were recommended for support services at discharge compared with the control group (OR = 1.63; 95% CI = 1.05–2.54; P = 0.03). Table 2 shows that this difference was due to a significantly higher proportion of the test group being recommended for home nursing (OR = 2.10; 95% CI = 1.29–3.41; P = 0.002). Table 2 also shows that there was no difference between the groups in terms of being discharged to supported accommodation (OR = 0.81; 95% CI = 0.52–1.26; P = 0.34).
|
Length of stay
Average days length of stay did not differ between the groups (F = 1.44; d.f. = 1; P = 0.23; test group 25 days, SD = 20; control group 22 days, SD = 16).
Patient satisfaction with discharge planning
The test group were significantly more likely to report that their discharge plan had been discussed with them by hospital staff during their admission (89% test; 69% controls; OR 5.01; 95% CI = 2.28–11.00; P < 0.0001). Eighty per cent of those in the test group who received a pre-discharge visit reported finding it useful.
At 6-week follow-up, significantly more of the test group reported that their return home was well prepared (93% versus 82%; OR = 2.72; 95% CI = 1.09–6.82; P = 0.03).
Readmission to hospital
Twenty-eight per cent (n = 101) of the sample experienced one or more unplanned readmissions up to 6 months after their initial discharge, but this did not differ between study groups (30% versus 25%; OR = 1.34; 95% CI = 0.83–2.17; P = 0.22).
Days to first readmission
No differences were detected between the groups overall, although analysis of the non-supported group showed that the test group experienced a non-significant greater mean number of days to first readmission (F = 1.56; d.f. = 1; P = 0.22; test group 60 days, SD = 58; control group 43 days, SD = 40).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This randomized controlled trial was conducted to evaluate the impact of GP pre-discharge visits on the discharge planning and post-discharge outcomes of frail elderly in-patients. Although outcome assessment was not blinded, potential bias was minimized as data such as readmissions, length to first readmission and community service referrals were verified through medical records. The frail nature of the sample (as indicated by the high readmission and death rates) limits generalizability of the results to similar patients in hospital geriatric units. Analysis was undertaken on the basis of group allocation, rather than considering only those patients who actually received a visit (52% of patients randomized to receive a GP visit), in order to avoid bias which might have resulted from differential quality of care provided by GPs who did or did not visit their patients in hospital.
That a significantly higher proportion of the intervention group patients reported that discharge plans had been discussed with them while in hospital, and that their return home was well prepared, is a significant benefit for this group of patients, who may feel excluded from decisions about post-discharge arrangements and may be apprehensive about returning home after a long period of hospitalization. Most patients receiving a GP visit reported that it was useful. In this sense, GPs may be playing an important advocacy role in terms of conveying, on their patient's behalf, important information to hospital teams about need for community services to maintain the patient in his/her own home following discharge.
Alternatively, it may be the case that those who receive a pre-discharge visit by a GP benefit from extra attention bestowed by hospital staff involved in arranging visits and liaising with the GP, or that the pre-discharge visit triggers a process which results in greater face-to-face communication opportunities between the hospital and GPs. Patients may also feel reassured by seeing someone who is familiar and knows their history. Examination of the aspects of GP visits felt to be useful by hospital in-patients is worthy of future study.
A higher proportion of patients in the intervention group were recommended for community services, particularly home nursing, and this effect persisted even when data were adjusted for Barthel's score. This indicates that GP pre-discharge visits can alter the level and type of community services implemented at discharge, and that specifically hospital staff acted on GPs' advice and capitalized on GP knowledge of their patients' likely coping ability and home circumstances. In addition, the trend for ‘test’ patients in the non-supported group to experience a longer time to first readmission may also reflect the greater community services received from this group, which may support the patient at home for a longer period.
An impact of GP pre-discharge input on ‘hard’ end-points such as length of stay and readmissions was not detected. This may reflect the increasing trend towards early commencement of discharge planning in geriatric units and also the multiple morbidity of this group of frail, elderly in-patients. It is possible that if the criteria were expanded to include all older people admitted under a geriatrician, the potential for pre-discharge visits to reduce readmissions and hospital costs may be realized. It may also be that the use of such outcomes, frequently demanded by health planners, may not be meaningful indicators of quality of care in frail elderly in-patients.
In conclusion, although involvement of the GP in discharge planning was associated with greater use of community resources, which may provide a less cost-effective discharge plan, this was compensated for by greater patient satisfaction and improved hospital–GP collaboration and communication opportunities. This may be an important trade-off, especially in frail aged care where it may not be possible to alter the likelihood of hard outcomes such as readmission rates.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 Balla JI, Jamieson WE. Improving the continuity of care between general practitioners and hospitals. Med J Australia1994; 161: 656–659.
2
Meara JR, Wood JL, Wilson MA, Hart MC. Home from hospital: a survey of hospital discharge arrangements in Northamptonshire. J Pub Health Med 1992; 14: 145–150.
3
Reeve H, Baxter K, Newton P, Burkey Y, Black M, Roland M. Long-term follow-up in outpatient clinics. 1: The view from general practice. Fam Pract 1997; 14: 24–28.
4 Isaac DR, Gijsbers AJ, Wyman KT, Martyres RF, Garrow BAC. The GP–hospital interface: attitudes of GPs to tertiary teaching hospitals. Med J Aust 1997; 166: 9–12.[Web of Science][Medline]
5 Ranmuthugala G, McInnes E, Mira M, Rendalls S, Atkin N, Kennedy P. A pre-discharge project—does willingness equal involvement? Aust Fam Physician 1997; 26 (suppl 2): S104–S108.
6
Schattner P, Dunt D. General Practitioner involvement in non-procedural medicine in public hospitals in Melbourne, Australia. Fam Pract 1989; 6: 141–145.
7 Montalto M, Gunesekera A, Sanderson F. The use of inpatient admission privileges by members of a Division of General Practice. Aust Fam Physician 1994; 3: 453–461.
8 Townsend J, Piper M, Frank AO, Dyer S, North WRD, Meade TW. Reduction in hospital readmission stay of elderly patients by a community based hospital discharge scheme: a randomized controlled trial. Br Med J 1988; 297: 544–547.
9 Weinberger M, Oddone EZ, Henderson WG. Does increased access to primary care reduce hospital readmissions? N Engl J Med 1996; 10: 1441–1447.
10 Harris RD, Hensche PJ, Popplewell PY et al. A randomized study of outcomes in a defined group of acutely ill elderly patients managed in a geriatric assessment unit or a general medical unit. Aust NZ J Med 1991; 21: 230–234.[Web of Science][Medline]
11 Siu AL, Kravitz RL, Keeler E et al. Postdischarge geriatric assessment of hospitalized frail elderly patients. Arch Intern Med 1995; 156: 15–17.
12 Wilkin D, Hallam L, Doggett MA. Measures of Need and Outcome for Primary Health Care. UK: Oxford University Press, 1994.
13 Tombaugh T, McIntyre N. The mini-mental status examination: a comprehensive review. J Am Geriatr Soc 1992; 40: 922–935.[Web of Science][Medline]
14 SPSS/PC+ Base System User's guide. Version 5. Chicago: SPSS Inc, 1992.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Kripalani, F. LeFevre, C. O. Phillips, M. V. Williams, P. Basaviah, and D. W. Baker Deficits in Communication and Information Transfer Between Hospital-Based and Primary Care Physicians: Implications for Patient Safety and Continuity of Care JAMA, February 28, 2007; 297(8): 831 - 841. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Schers, C. van de Ven, H. van den Hoogen, R. Grol, and W. van den Bosch Patients' Needs for Contact With Their GP at the Time of Hospital Admission and Other Life Events: A Quantitative and Qualitative Exploration Ann. Fam. Med, September 1, 2004; 2(5): 462 - 468. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. K Mitchell How well do general practitioners deliver palliative care? A systematic review Palliative Medicine, September 1, 2002; 16(6): 457 - 464. [Abstract] [PDF]
|
|
|
|
 |
|
 S. A. Walker and J. M. Eagles Discharging psychiatric patients from hospital Psychiatr. Bull., July 1, 2002; 26(7): 241 - 242. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||