Review of preventive home visits to elderly people: Is vote counting the answer?
EDITOR - Haastregt et al conclude that there is little evidence
supporting the effectiveness of preventive home visits to elderly people
living in the community (1). Their review is timely but methodological
shortcomings limit its usefulness. The principal method of analysis
consisted of "vote counting": adding up the number of studies showing
statistically significant effects. This procedure is a sad relic from the
times of unsystematic, narrative reviews, ignoring sample size, effect
size, type of intervention and methodological quality (2). For example, it
is important to distinguish between preventive home visits that included
multidimensional geriatric assessment with follow-up and interventions
that did not (3).
Assessment of the quality of trials was also problematic. Empirical
research has shown that the scale used by Haastregt et al, and scales in
general, may produce misleading results (4). Rather than calculating a
summary score the methodological aspects that are important in a given
context should be identified and assessed individually. Some of the items
included by Haastregt et al are not relevant in this context ("co-
interventions" are an integral part of preventive home visits), others do
not measure the quality of a trial (adverse effects are an important
outcome but not a measure of methodological quality), while important
dimensions of quality (for example, concealment of allocation) were
lacking from the list of items.
Three of us (AS, JCB, CM) were involved in a randomised trial of
preventive home visits which was conducted in Berne, Switzerland (5). The
findings from this trial, which was published after the review by
Haastregt et al appeared showed that preventive home visits can reduce
disability, which in a 3-year period may save up to 1400 US $ per person
per year (5). In a planned subgroup analysis we found that the effect of
the intervention depended on the base-line risk status of trial
participants (disability was reduced among persons at low-risk at base-
line, but not among high-risk participants). In addition, the professional
experience of the person visiting was an important factor determining
programme efficacy. These findings indicate that the composition of the
study population and the type and quality of the intervention are
important factors which may explain the discrepant results obtained from
randomised trials of preventive home visits.
Although there are conflicting results between individual trials of
preventive home visits, some trials clearly demonstrate that home visits
can substantially reduce or delay the onset of disability. Thus, research
is needed to explicitly define the conditions for cost-effective
programmes for reducing disability among older people. We agree with
Haastregt et al that it is often inappropriate to combine a heterogeneous
set of trials. However, vote counting cannot identify the factors
introducing heterogeneity. Further meta-analytic and trial research is
required to clarify what components of this complex intervention work in
which population groups.
Andreas Stuck, medical director
Department of Geriatrics and Rehabilitation, 3001 Bern, Switzerland
Matthias Egger, senior lecturer in epidemiology and public health
medicine
MRC Health Services Research Collaboration,
Department of Social Medicine, University of Bristol
Christoph E. Minder, senior biostatistician
Department of Social and Preventive Medicine, University of Bern
Steve Iliffe, reader in general practice
Department of Primary Care & Population Sciences, Royal Free &
University College
London Medical School
John C. Beck, MD, professor of medicine (geriatrics),
UCLA School of Medicine, Los Angeles, USA
References
(1) Van Haastregt JCM, Diederiks JPM, van Rossum et al. Effects of
preventive home visits to elderly people living in the community:
systematic review. BMJ 2000;320:754-8.
(2) Egger M, Davey Smith G. Rationale, potentials and promise. In:
Egger M, Davey Smith G, Altman DG. Systematic Reviews in Health Care: Meta
-Analysis in Context. London: BMJ Books (in press).
(3) Stuck AE, Walthert J, Nikolaus T, Büla CJ, Hohmann C, Beck JC.
Risk factors for functional status decline in community-dwelling elderly
people: a systematic literature review. Soc Sci Med 1999; 48:445-469.
(4) Jüni P, Witschi A, Bloch R, Egger M. The hazards of scoring the
quality of clinical trials for meta-analysis. JAMA 1999; 282:1054-1060
(5) Stuck AE, Minder CE, Peter-Wüest I, Gillmann G, Egli C,
Kesselring A, Leu RE, Beck JC. A randomized trial of in-home visits for
disability prevention in community-dwelling older people at low and at
high risk for nursing home admission. Arch Intern Med 2000 . Arch Intern
Med 2000; 160:977-986
Rapid Response:
Review of preventive home visits to elderly people: Is vote counting the answer?
EDITOR - Haastregt et al conclude that there is little evidence
supporting the effectiveness of preventive home visits to elderly people
living in the community (1). Their review is timely but methodological
shortcomings limit its usefulness. The principal method of analysis
consisted of "vote counting": adding up the number of studies showing
statistically significant effects. This procedure is a sad relic from the
times of unsystematic, narrative reviews, ignoring sample size, effect
size, type of intervention and methodological quality (2). For example, it
is important to distinguish between preventive home visits that included
multidimensional geriatric assessment with follow-up and interventions
that did not (3).
Assessment of the quality of trials was also problematic. Empirical
research has shown that the scale used by Haastregt et al, and scales in
general, may produce misleading results (4). Rather than calculating a
summary score the methodological aspects that are important in a given
context should be identified and assessed individually. Some of the items
included by Haastregt et al are not relevant in this context ("co-
interventions" are an integral part of preventive home visits), others do
not measure the quality of a trial (adverse effects are an important
outcome but not a measure of methodological quality), while important
dimensions of quality (for example, concealment of allocation) were
lacking from the list of items.
Three of us (AS, JCB, CM) were involved in a randomised trial of
preventive home visits which was conducted in Berne, Switzerland (5). The
findings from this trial, which was published after the review by
Haastregt et al appeared showed that preventive home visits can reduce
disability, which in a 3-year period may save up to 1400 US $ per person
per year (5). In a planned subgroup analysis we found that the effect of
the intervention depended on the base-line risk status of trial
participants (disability was reduced among persons at low-risk at base-
line, but not among high-risk participants). In addition, the professional
experience of the person visiting was an important factor determining
programme efficacy. These findings indicate that the composition of the
study population and the type and quality of the intervention are
important factors which may explain the discrepant results obtained from
randomised trials of preventive home visits.
Although there are conflicting results between individual trials of
preventive home visits, some trials clearly demonstrate that home visits
can substantially reduce or delay the onset of disability. Thus, research
is needed to explicitly define the conditions for cost-effective
programmes for reducing disability among older people. We agree with
Haastregt et al that it is often inappropriate to combine a heterogeneous
set of trials. However, vote counting cannot identify the factors
introducing heterogeneity. Further meta-analytic and trial research is
required to clarify what components of this complex intervention work in
which population groups.
Andreas Stuck, medical director
Department of Geriatrics and Rehabilitation, 3001 Bern, Switzerland
andreas.stuck@zieglerspital.bern
Matthias Egger, senior lecturer in epidemiology and public health
medicine
MRC Health Services Research Collaboration,
Department of Social Medicine, University of Bristol
Christoph E. Minder, senior biostatistician
Department of Social and Preventive Medicine, University of Bern
Steve Iliffe, reader in general practice
Department of Primary Care & Population Sciences, Royal Free &
University College
London Medical School
John C. Beck, MD, professor of medicine (geriatrics),
UCLA School of Medicine, Los Angeles, USA
References
(1) Van Haastregt JCM, Diederiks JPM, van Rossum et al. Effects of
preventive home visits to elderly people living in the community:
systematic review. BMJ 2000;320:754-8.
(2) Egger M, Davey Smith G. Rationale, potentials and promise. In:
Egger M, Davey Smith G, Altman DG. Systematic Reviews in Health Care: Meta
-Analysis in Context. London: BMJ Books (in press).
(3) Stuck AE, Walthert J, Nikolaus T, Büla CJ, Hohmann C, Beck JC.
Risk factors for functional status decline in community-dwelling elderly
people: a systematic literature review. Soc Sci Med 1999; 48:445-469.
(4) Jüni P, Witschi A, Bloch R, Egger M. The hazards of scoring the
quality of clinical trials for meta-analysis. JAMA 1999; 282:1054-1060
(5) Stuck AE, Minder CE, Peter-Wüest I, Gillmann G, Egli C,
Kesselring A, Leu RE, Beck JC. A randomized trial of in-home visits for
disability prevention in community-dwelling older people at low and at
high risk for nursing home admission. Arch Intern Med 2000 . Arch Intern
Med 2000; 160:977-986
Competing interests: No competing interests