Family Practice Vol. 21, No. 2, 173-179
Family Practice Vol. 21, No. 2 © Oxford University Press 2004, all rights reserved.
Article |
Six month outcomes after emergency admission of elderly patients to a community or a district general hospital
East Devon PCT, a North and East Devon Health Authority, b Cornwall and Isles of Scilly Health Authority, c Budleigh Salterton, d Exeter & District Community Trust, e Royal Devon & Exeter Hospitals Trust and f University of Bristol, UK
E-mail: Ali.Round{at}EastDevon-PCJ.nhs.uk
Received 20 June 2003; Revised 22 October 2003; Accepted 3 November 2003.
Round A, Crabb T, Buckingham K, Mejzner R, Pearce V, Ayres R, Weeks C and Hamilton W. Six month outcomes after emergency admission of elderly patients to a community or a district general hospital. Family Practice 2004; 21: 173–179.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAppendix 1: excluded conditionsReferences |
|---|
Background. Emergency admissions account for 40% of National Health Service bed usage. Recent policy is to increase the role of intermediate care, which includes the use of community hospitals (CHs). However, the proposed expansion presumes that CH care is as effective as acute hospital care. No direct comparison of outcomes between CHs and district general hospitals (DGHs) has been undertaken.
Objectives. The aim of this study was to compare patient-based outcomes at 6 months following emergency admission to a DGH or CH.
Methods. We designed a prospective cohort study, with strict eligibility criteria. The study was set in one DGH and five CHs in Devon, UK. Study participants were people aged >70 years with an acute illness requiring hospital admission, but whose condition could have been treated in either hospital setting. A cohort of people admitted to each setting was identified and followed-up for 6 months. The primary outcome measure was change in quality of life 6 months after admission, as measured by SF-36 and EuroQol. Secondary outcome measures were death, readmission and place of residence at 6 months. The use of drugs and investigations during the hospital stay were also measured.
Results. A total of 376 patients were recruited and completed baseline measures, 254 of whom were followed-up at the 6-month stage (136 CH, 118 DGH). There were no differences in outcome between settings, with a small increase in quality of life scores at 6 months in both cohorts: the mean change in EuroQol 5D in CH was 6.6 points (95% confidence interval, 2.8–10.4) and in DGH was 6.5 (2.4–10.7); P = 0.97. Mortality and place of residence at 6 months were similar in the two groups. The numbers of investigations (median CH four investigations, DGH 22; P < 0.001) and of prescribed medications during the hospital stay (median CH eight drugs, DGH 11; P < 0.001) were significantly higher in the DGH.
Conclusions. The quality of life and mortality in the CH cohort was similar to those in the DGH cohort. CH care can be used as an alternative to DGH care for a wide range of conditions requiring emergency admission.
Keywords. Community hospital, emergency admission, quality of life.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1: excluded conditionsReferences |
|---|
Emergency hospital admissions account for 40% of NHS bed usage, and are increasing by 2–3% each year.1,2 Most of these admissions are to district general hospitals (DGHs) but in some parts of the UK there are also community hospitals (CHs). In these, the medical input is usually from GPs (as opposed to hospital consultant teams), there is less access to intensive therapy and investigation, and variable input from nurses and professions allied to medicine. There are nearly 500 such hospitals in the UK, with almost 20 000 beds. A similar model exists in many parts of the world.3 Around 20% of UK GPs have admitting rights to a CH.4 CHs are popular with patients5 and are usually situated closer to the patient's home than the DGH. This is a particular advantage for the elderly, especially in terms of visits by friends or relatives, who will often be elderly themselves.
One response to the increasing number of emergency admissions is to promote use of intermediate care. Intermediate care is described in the National Service Framework for Older people as services that will improve recovery and rehabilitation, increase independent living and reduce acute hospital admissions.6 This setting of care includes CHs. The Department of Health's Winter report for 2001 stated that, "[it wishes to] ensure that there is adequate access to appropriate care for older people in all parts of the country." However, the projected expansion of intermediate care presumes that it as effective as acute hospital care for appropriate cases.7 There is some research support for this view. Studies comparing more intensive care or rehabilitation against standard care, or specialist physicians as opposed to generalists within a single hospital, have not found any difference in outcomes.8–11 Likewise, evaluation of hospital at home schemes has revealed no difference in patient outcomes.12 However, little rigorous research has compared CHs with the alternatives available. One study of an urban GP unit with 15 beds found equivalent outcomes for patients admitted to their unit compared with conventional care; however, only a minority of those receiving conventional care had required a hospital admission.5
We therefore designed a study to measure clinically relevant outcomes of emergency admission in the elderly, comparing the two settings, CH and DGH.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1: excluded conditionsReferences |
|---|
We studied patients aged
70 years, requiring acute hospital care and admitted to a CH or DGH between November 1999 and November 2000. Patients in the DGH were recruited from the urban catchment area of the Royal Devon & Exeter Hospital NHS Trust. No CH is available for this population. CH patients were recruited from five East Devon CHs in the rural catchment area of the same DGH. The ideal study design would have been for rural GPs to randomize eligible patients to either setting. However, we considered it unethical to admit an elderly ill patient to a hospital up to 20 miles away from their home when one was available nearby. Thus we designed this prospective cohort study with strict inclusion and exclusion criteria, with the aim of producing two comparable groups. A pilot study was conducted to ascertain feasibility, determine sample size and derive eligibility criteria. The local research ethics committee approved the study and all patients recruited gave informed consent. All six hospitals were visited by a single researcher (TC) on a rotating basis to ensure equal coverage of all days of the week. Patients admitted within the previous 48 h to a CH, or to acute medical, emergency medical or geriatric wards in the DGH were potentially eligible. For inclusion, the patients had to have been admitted with an acute illness whose condition could have been fully treated in either type of hospital. Exclusion criteria were admissions that required specific hospital admission, such as gastrointestinal bleeding for the acute hospital or palliative care for the CHs (these are outlined in the Appendix). The nursing staff also excluded patients if they were too ill or confused to talk to the researcher.
Quality of life was measured at baseline using the SF-3613 and EuroQol 5D weighted health index and self-rated score.14 These are generic questionnaires, designed to be used across a wide range of health interventions, and have been validated for use in older people. Potential confounding or explanatory variables were collected at baseline using standard and validated questionnaires: Activities of Daily Living (ADL),15 Mini-mental State Examination (MMSE)16 and co-morbidity index.17 The initial diagnosis was recorded from the written hospital notes, and was grouped into broad categories by one researcher (AR) blinded to the hospital setting. All questionnaires were administered by a single researcher (TC). Patient satisfaction is reported separately.
Outcome measures
The primary outcome measure was change in quality of life (SF-36 and EuroQol) 6 months after the date of admission. Questionnaires were sent by post, and non-respondents contacted by telephone 2 weeks later. In this telephone call, patients were offered either a home visit or a duplicate questionnaire. Secondary outcome measures were mortality within 6 months, numbers of readmissions, length of stay and place of residence at 6 months. Dates of death were ascertained from computerized local registrars (death certification) and the patient administration systems of the hospitals. Readmissions, length of stay and place of residence at 6 months were collected from patient administration systems or by direct contact. The number of different drug prescriptions and the number of investigations performed during the index admission were recorded. For blood tests, multichannel investigations (e.g. full blood count) counted as one investigation. All these measures were selected pre hoc.
Sample size
This was determined for the primary outcome measure, a difference in change in SF-36 scores between baseline and 6 months, using the pilot data. A sample size of 119 in each group was sufficient for a difference between settings of 5 points, with 90% power at a significance level of 0.05, and assuming an SD of 15 points.18 Using an estimated drop-out rate of 30% following baseline data collection, we aimed to recruit 350 people in total.
Statistical analysis
We compared the change in quality of life at 6 months using a t-test for SF-36 overall and subscale scores, and EuroQol self-rated scores, as these were normally distributed. EuroQol weighted health index scores were compared by a Mann–Whitney U-test, as the distribution was skewed. Multivariable regression was used to adjust for the effects of covariables. Sex and age were included as covariables in all multivariable analyses; other baseline variables were included if the univariable significance was <0.1. Patients were included in their original group irrespective of any subsequent treatment received.
Death was reported as an overall proportion. A log rank test using date of death was performed to compare survival between hospital settings. A two-sample test of proportions was used for comparing the numbers of readmissions and place of residence at 6 months. Length of stay, and the numbers of drugs and investigations were compared using a Mann–Whitney U-test. Comparison of categorical variables was by chi-squared test or Fischer's exact test. All analyses were performed in STATA, version 7.19
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1: excluded conditionsReferences |
|---|
Between June 1999 and November 2000, a total of 376 patients were recruited and completed baseline measures, 254 of whom were followed-up at the 6-month stage. Figure 1 shows the trial profile.
|
The proportion of patients who were too ill or declined consent to participate was the same in both groups (53% CH versus 59% DGH, chi-squared 1.05, P = 0.31). The numbers of deaths and withdrawals prior to completing the first baseline measure was also similar in both cohorts (23% CH versus 19% DGH, chi-squared 0.57, P = 0.45). Deaths and withdrawals between baseline and outcome were 34% in CH and 31% in DGH (chi-squared 2.5, P = 0.51).
Comparability between the cohorts
The two cohorts were similar in terms of baseline characteristics, except that the CH cohort were older and had a higher proportion of females (Table 1). Those people who completed the outcome measurements were more likely to be male and had a slightly higher quality of life compared with all those who completed the baseline measures [54.4% male versus 52.5%, difference 1.9, 95% confidence interval (CI) -0.2 to 4.1]. Overall, both cohorts were in poor health with substantial co-morbidity and moderately impaired functional ability. The diagnoses on admission are shown in Table 2.
|
|
Primary outcome measure—quality of life
Results from these are shown in Table 3. In none of the measures was there a difference between hospital settings. There was a small and significant increase in quality of life at 6 months on the EuroQol 5D self-rated score and EuroQol 5D weighted health index compared with baseline for both cohorts. There was no change over time on the overall SF-36 score, but the subscales were significantly improved for physical functioning and all the well-being scales. The general health perception, social functioning and role limitations (emotional problems) were all significantly worse, but again with no difference between the two cohorts.
|
Multivariable regression confirmed that hospital setting was not a determinant of change in quality of life at 6 months. A higher baseline quality of life was associated with a smaller improvement. Conversely, greater functional ability, as measured on the ADL scale, was associated with a larger improvement (Table 4). Age was not a determinant of improvement in quality of life, except on the social and physical functioning subscales of the SF-36 (data not shown).
|
Secondary outcome measures
Twenty-two (18.8%, 95% CI 12.9–24.7) of the DGH cohort and 23 (17.0%, CI 11.9–22.1) of the CH cohort had died within 6 months, with a hazard ratio of 0.85 (CI 0.53–1.34) for admission to the CH (log rank test, chi-squared = 1.2, P = 0.48). At the 6 month assessment, three people (2.5%, CI 0.5–7.3) from the DGH cohort and 10 (7.4%, CI 3.6–13.1) from the CH cohort were living in a nursing or residential home (P = 0.76). There were 17 readmissions to any hospital in the DGH cohort (14.4%, CI 8.1–20.7) and 15 (11%, CI 5.7–16.3) in the CH cohort. This is not significant (P = 0.41). Median length of stay was no different, being 9 days [interquartile range (IQR) 6–17] in the DGH cohort and 11 days (IQR 5–20.5) in the CH cohort (Mann–Whitney U-test, P = 0.83).
There was a significantly higher level of intervention in the DGH cohort, for both prescribed drugs and investigations (Table 5).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1: excluded conditionsReferences |
|---|
Main findings
This study is the first to compare outcomes following emergency admission to a CH or DGH. There were no significant differences between settings in quality of life at 6 months. This is an important finding, and supports the current provision of CH care and promotion of intermediate care as an alternative to acute hospital care.
The patients in this study mirror the sort of management problem seen regularly in primary care. They were patients with poor underlying health, who developed an additional problem such as an infection, making hospital admission necessary. The GP who can admit to either a CH or a DGH has to decide which facility is better for that patient. The improvement in quality of life scores was small (but equally so) in both groups. This might be expected, given the age, functional status and co-morbidity of the cohorts. Our results agree with other research showing small changes in quality of life after hospital admission.20 Emergency hospital admission appears to have a limited impact in this frail group, whose baseline quality of life is relatively low. There were also no differences in mortality, readmission rates and place of residence, although the study was not powered to identify these. Concerns that care outside a DGH may result in increased numbers of readmissions20 are not supported by our results, but we cannot exclude a small increase, given the wide confidence intervals.
Study limitations
As the study was not randomized, it is possible that the two groups were not comparable, despite the strict eligibility criteria. The baseline measurements were similar in the two groups and, as the functional status has been shown to be the major determinant of outcome,21–23 this is probably more important than the relatively small age and sex difference between the groups. The diagnoses recorded on admission showed some differences. This may partly reflect the study design, which did not have specific diagnostic inclusion criteria, but excluded conditions that would be treated preferentially in one site or another. It may also be partly artefactual. GPs in the CH had previous knowledge of the patient, and were admitting to their own beds. For an admission to the DGH, which will have to be accepted by the on-call team, it is quite possible that a more formal diagnostic label was given to ensure that admission was accepted. It is unlikely that the differences in diagnostic labelling are a major source of bias, given the similar functional status between the two groups.
As expected with such ill patients, there were high numbers of patients excluded, withdrawals and deaths. Those who completed the outcome assessment had a higher quality of life than those who competed the baseline assessment, probably as a result of selective withdrawal and death of the more severely ill. However, there was a similar pattern for these in both cohorts, and significant bias seems unlikely.
The GPs and social services were different for the two cohorts. It is possible that differences in health and social care support subsequent to the hospital admission have affected the final outcomes. Although there are no apparent differences, we did not examine these services in detail. It would, however, be surprising if one of the hospital settings had an inferior record in hospital treatment of their patients, which was then exactly cancelled out by superior community care, giving our overall equal result.
Implications of the study
Policymakers need to know if the proposed boost to intermediate care is appropriate: GPs need to know which hospital to admit their patients to. These points alone make research on CHs important. However, research in this field is not easy. Pragmatic decisions have to be made when studying ill, aged people. Nonetheless, the fact that research is difficult, and that conclusions have to be expressed cautiously, does not reduce the need to study the subject.
Given the findings of equivalent outcomes 6 months after admission, both policymakers and GPs can be reassured. For the patient who could be admitted to either facility, neither setting was superior. It must be remembered that these patients comprise the minority of admissions, as shown by the proportion of patients ineligible for the study. For the policymaker, new research questions arise.
Further research is needed to confirm, or refute, the mortality findings. The economic consequence of care in alternative settings needs to be explored in detail. Our data suggested that medical staff contact and use of drugs and investigations were major sources of cost differences between the cohorts. Exploration of the benefits and disadvantages of multiple investigations and drug treatments in a frail elderly cohort would also be worthwhile.
Finally, the training and staffing implications of an expansion in intermediate care have not been well studied. It is uncertain how many GPs are willing and skilled to take on this extra role, or whether, given the current shortage of GPs, this would be an incentive or barrier to recruitment.
|
Appendix 1: excluded conditions |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1: excluded conditionsReferences |
|---|
- Pneumothorax
- Meningitis
- Arrhythmia
- Possible deep vein thrombosis
- Abnormal electrolytes/calcium
- Possible pulmonary embolus
- Subarachnoid haemorrhage
- Renal failure
- Acute chest pain
- Gastrointestinal bleed
- Subspecialty admissions
- Diabetic ketoacidosis
- Overdose
- Severe cognitive impairment
- Terminal care
- Myocardial infarction
- Endocarditis
- Liver failure
|
Acknowledgments |
|---|
We thank the patients, clinicians and managers involved in this study for their help. Hilary Parkman, Cindy Baker and the Barnfield Hill Research Practice helped with data collection. The study was funded by a grant from the Charitable Trustees of North & East Devon Health Authority.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1: excluded conditionsReferences |
|---|
1 Capewell S. The continuing rise in emergency admissions. Br Med J 1996; 312: 991.
2 Pettinger N. Emergency admissions. Change the record. Health Serv J 2001; 111: 20–21.[Medline]
3 Lemieux-Charles L, Leatt P. Hospital-physician integration: case studies of community hospitals. Health Serv Manag Res 1992; 5: 82–98.
4 Seamark D, Moore B, Tucker H, Church J, Seamark C. Community hospitals for the new millennium. Br J Gen Pract 2001; 51: 125–127.[Web of Science][Medline]
5 Boston NK, Boynton PM, Hood S. An inner city GP unit versus conventional care for elderly patients: prospective comparison of health functioning, use of services and patient satisfaction. Fam Pract 2001; 18: 141–148.
6 Department of Health. The National Service Framework for Older People. London: Department of Health; 2001.
7 Hensher M, Fulop N, Coast J, Jefferys E. Better out than in? Alternatives to acute hospital care. Br Med J 1999; 319: 1127–1130.
8 Coleman EA, Kramer AM, Kowalsky JC et al. A comparison of functional outcomes after hip fracture in group/staff HMOs and fee for service systems. Effective Clin Pract 2000; 3: 229–239.
9 Asplund K, Gustafson Y, Jacobsson C et al. Geriatric based versus general wards for older acute medical patients: a randomized comparison of outcomes and use of resources. J Am Geriatr Soc 2000; 11: 1381–1388.
10 McGann KP, Bowman MA, Davis SW. Morbidity, mortality, and charges for hospital care of the elderly: a comparison of internist's and family physicians' admissions. J Fam Pract 1995; 40: 443–448.[Web of Science][Medline]
11 McVey LJ, Becker PM, Slatz CC, Feussner JR, Cohen HJ. Effect of a geriatric consultation on functional status of elderly hospitalised patients. A randomized controlled clinical trial. Ann Intern Med 1989; 110: 79–84.
12 Shepperd S, Iliffe S. Hospital at home compared to inpatient care (Cochrane review). In: The Cochrane Library. 2nd edn: Update Software, 2002.
13 Walters SJ, Munro JF, Brazier JE. Using the SF-36 with older adults: a cross-sectional community based survey. Age Ageing 2001; 30: 337–343.
14 Kind P, Gudex C, Dolan P, Williams A. Practical and Methodological Issues in the Development of the EuroQol. The York Experience. J.A.I. Press; 1994.
15 Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW. Studies of illness in the aged. The Index of ADL: a standardized measure of biological and psychosocial function. J Am Med Assoc 1963; 185: 914–919.
16 Folstein MF, Folstein SE, McHugh PR. Mini-mental state: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: 189–198.[CrossRef][Web of Science][Medline]
17 Incalzi RA, Capparella O, Gemma A et al. The interaction between age and comorbidity contributes to predicting the mortality of geriatric patients in the acute care hospital. J Intern Med 1997; 242: 291–298.[CrossRef][Web of Science][Medline]
18 Machin D, Campbell M, Fayers P, Pinol A. Sample Size Tables for Clinical Studies, 2nd edn. Oxford: Blackwell Science; 1997.
19 Stata Statistical Software: Release 7.0 [program]. College Station (TX): Stata Corporation; 2001.
20 Shepperd S, Harwood D, Jenkinson C, Gray A, Vessy M, Morgan P. Randomised controlled trial comparing hospital at home care with inpatient hospital care. I: three month follow up of health outcomes. Br Med J 1998; 316: 1786–1791.
21 Alarcon T, Barcena A, Gonzalez-Montavo JI, Penalosa C, Salgado A. Factors predictive of outcome on admission to an acute geriatric ward. Age Ageing 1999; 27: 429–432.
22 Caplan GA, Brown A, Croker WD, Doolan J. Risk of admission within 4 weeks of discharge of elderly patients from the emergency department. the DEED study. Discharge of elderly from emergency department. Age Ageing 1999; 27: 697–702.
23 Inouye SK, Peduzzi PN, Robison JT, Hughes JS, Horwitz RI, Concato J. Importance of functional measures in predicting mortality among older hospitalized patients. J Am Med Assoc 1998; 279: 1187–1193.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Garasen, R. Windspoll, and R. Johnsen Long-term patients' outcomes after intermediate care at a community hospital for elderly patients: 12-month follow-up of a randomized controlled trial Scand J Public Health, March 1, 2008; 36(2): 197 - 204. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||