A randomised controlled trial

ORIGINAL PAPER

æ

Structured home visits to older people. Are they only of benefit for women? A randomised controlled trial Mikkel Vass1, Kirsten Avlund2, Kajsa Kvist3, Carsten Hendriksen2, Christian K. Andersen4 and Niels Keiding3 1

Scand J Prim Health Care Downloaded from informahealthcare.com by 14.37.69.97 on 05/20/14 For personal use only.

Department of General Practice and Central Research Unit for General Practice, Institute of Public Health, University of Copenhagen, 2Department of Social Medicine, Institute of Public Health, University of Copenhagen, 3Department of Biostatistics, Institute of Public Health, University of Copenhagen, 4Health Economics Research Unit, University of Southern Denmark, Odense, Denmark.

Scand J Prim Health Care 2004;22:106 /111. ISSN 0281-3432

Objective / To investigate whether education of primary care professionals improved functional ability in home-dwelling older people, with special focus on gender differences. Design / A prospective controlled three-year follow-up study (1999 / 2001) with randomisation and intervention at municipality level and outcomes measured at individual level . Intervention municipality visitors received regular education and GPs were introduced to a short assessment programme. Control municipalities received no education but conducted the preventive programme in their own way. Setting / Primary care, 34 municipalities. Subjects / 5788 home-dwelling 75- and 80-year-olds were invited. 4060 (70.1%) participated: 2104 in 17 intervention- and 1956 in 17 matched control-municipalities. The main outcome measure was obtained from 3383 (95.6%) of 3540 surviving participants. Main outcome measure / Functional ability.

The beneficial effect of in-home assessment of elderly people on disability prevention or delay claimed in a number of randomised trials over the past two decades remains controversial (1 /4). The conflicting nature of these results may be ascribed to methodological differences, notably in which population to target, the intensity of the interventions, and to differences in the primary care systems. Uncertainty centres on the feasibility and cost-effectiveness of educational programmes in routine primary care. Men and women may differ in adherence to life style recommendations (5). Patterns of functional decline vary for men and women (6). No studies have considered that the influence of preventive home visits on functional decline may have gender differences. It is claimed that preventive home visits cannot be assessed by means of randomised controlled trials as such services comprise a complex mix of uncontrollable, independent variables embedded in what is a social process more than a treatment programme (7,8). However, it seems relevant to argue that good health and functional ability is a robust outcome that may embrace both the individual and the medical/administrative discourse. Scand J Prim Health Care 2004; 22

Results / Municipality intervention in coordination with GPs was associated with better functional ability in women (OR: 1.26; CI95: 1.08 /1.47, p/0.004), but not in men (OR: 1.04; CI95: 1.85 /1.27). Accepting and receiving free preventive home visits was associated with better functional ability among women (OR: 1.36; CI95: 1.16 /1.60, p/0.0002), but not among men (OR: 0.98; CI95: 0.80 /1.21). Conclusion / A brief, feasible educational intervention for primary care professionals and to accept and receive preventive home visits may have effect in older women, but not in older men.

Key words: older people, preventive home visits, assessment, functional ability, community intervention, longitudinal studies. Mikkel Vass, Department of General Practice, Institute of Public Health, University of Copenhagen, Blegdamsvej 3, DK-2200 Copenhagen, Denmark, E-mail: [email protected]

Preventive home visitation programmes in elderly people have been part of national policy in the UK since 1990 and in Australia since 1999 (4,9). All Danish municipalities have been required by law since 1998 to offer two annual preventive home visits to all citizens aged 75 years or older. After the law had been in force for a few years, many municipalities had come to recognize a need for more knowledge of how best to carry out the programme. Visits today are primarily

Preventive home visits to elderly people remains controversial. Uncertainty centres on feasibility and cost-effectiveness in routine primary care. . This study suggests that education of home visitors and GPs will prevent functional decline in home-dwelling older women, but not in older men. . Accepting and receiving preventive home visits will prevent functional decline in homedwelling older women, but not in older men.

DOI 10.1080/02813430410005829

Structured home visits to older people


carried out by district nurses, but an obligatory health check is not included, and the general practitioners (GPs) are rarely directly involved. This study aimed to investigate whether education of home visitors and GPs in routine primary care improved the functional ability in home-dwelling older people, with special focus on gender differences. MATERIAL AND METHODS The study was designed as a controlled three-year follow-up study (1999 /2001) with randomisation and intervention at municipality level and outcome measurement at individual level. Detailed information about the study design has been published elsewhere (10). Municipality inclusion criteria The municipality should offer preventive home visits as prescribed by law. Possibility of GP participation and structural facilitation of good rehabilitation should be present. Fifty of 81 municipalities in four counties met these criteria and were invited. 34 municipalities agreed to participate. No demographic differences were seen between these and the remaining 16 municipalities. Randomisation Randomisation was carried out following paired matching of intra-county municipalities, urban/rural type, size and geriatric services. No major differences between intervention and control municipalities were seen (10). Study population Altogether, 5788 non-institutionalized citizens born in 1918 or 1923/1924 received a postal letter of invitation. In the large municipalities, a random sample or a community was chosen. Addresses were drawn from the Civil Registration Office. We obtained written consent from 4060 persons (participation rate 70.1%) and there were no major differences in baseline characteristics between intervention and control participants (10). Cognitively impaired participants were not actively excluded. Twenty-two persons died and four were institutionalised before the intervention started, leaving 4034 persons in the study population. Intervention We encouraged conducting the home visits in a structured way focusing on early signs of disability and on physical activity while respecting individual variation, and endeavouring inter-disciplinary coordinated follow-up in the local setting (11 /18). The contents of the intervention (Table I) were based on

107

updated geriatric and gerontological documentation. Twice a year we introduced a standard assessment tool to two key persons from each of the intervention municipalities, promoted training in its use and interpretation, encouraged these key persons to use the tool themselves and to train others in the use. Tiredness in daily activities should be interpreted as an early sign of disability and alert the visitor to search for the reason for such tiredness in the health, mental or social domains (13). Any suspicion of a health problem would result in contact to the GP, who was expected and urged to take the encounter seriously and to incorporate a short geriatric assessment in his/ her usual clinical practice. Outcome measure Functional ability was measured at baseline, at 18month and three-year follow-up. The survivors’ functional decline was assessed using a validated mobility scale (the Mob-H Scale) which included items about need for help in transferring, walking indoors, going outdoors, walking outdoors in nice weather, walking outdoors in poor weather, and climbing stairs (6,19). The scale was included in the analyses as a dichotomized variable: manage all activities without help versus need of help in one or more activities. Covariates Covariates were: Number of home visits (receiving no versus receiving at least one home visit during the three years), sex (specified from the Civil Registration Office), age (born in 1918 or 1923/24), and time (18 months and three years after baseline). The 17 pairs of municipalities were based on the matched randomisation (1 to17), and live alone measured with answer ‘yes’ and ‘no’ by a question at baseline. Ethics The regional Ethical Committee involved approved the study. Statistical design considerations All analyses were stratified by sex. The study of longitudinal data in aging research requires due consideration of participants who disappear from view either because of drop-out or death. In survival and more general event history analysis, it is customary to consider such drop-outs as censorings, and the standard approach is essentially to see the remaining study individuals as representatives of the drop-outs. This approach seems justified here for those dropping out alive, particularly under the assumption that they have the same survival potential as those remaining in the study. The standard procedure would use this approach also for those who died, assuming that death Scand J Prim Health Care 2004; 22

108

M. Vass et al.

Table I. The education for primary care professionals. The intervention 1. 2. 3. 4.

Initial interdisciplinary education of all professionals involved in preventive home visits Education of two key persons from each community followed up twice a year Introduction of an implementation strategy to prevent falls Small group-based education of the GPs The main content of the education given to the key persons

/ To emphasize the importance of psychological, social as well as health factors in the assessment, underlining the health

dimension, but no specific physical or mental examination was recommended. / To focus on early signs of disability, especially unexplained tiredness in daily activities. Two easily administered functional

ability tests were recommended (13,15). Scand J Prim Health Care Downloaded from informahealthcare.com by 14.37.69.97 on 05/20/14 For personal use only.

/ To stress the importance of physical activity and stimulate the communities to facilitate participation in physical activities

through convenient transportation and sports for elderly people (17). / To focus on relevant geriatric problems, e.g., prevention of falls, mental problems, medication, incontinence, and nutrition

(16) and refer to the GP when appropriate. The main message to the GPs was / Take any encounter caused by a preventive home visit seriously / Think twice before you say ‘it is age’ / Incorporate the 5 D’s in your usual clinical practice:

Disease, Depression, Dementia, Drugs, Drinks

is a part of a functional decline pattern, and we would then essentially extrapolate functional ability beyond death. However, an important recent methodological contribution demonstrated how to avoid this extrapolation (20). It restricts attention to the functional ability of the survivors while still considering live dropouts as censorings, technically using inverse probability weighting. The described method was employed to account for the missing values and the repeated measurements were handled using generalized estimating equations. The Mob-H scores were dichotomized due to software limitations and analyzed using logistic regression. The mortality of the cohort was investigated using a Cox regression model. Analyses were based on the intention to treat principle and SAS software.

RESULTS During the three study years, 1198 (57%) of the participants accepted and received at least one preventive home visit in the intervention municipalities compared with 1184 (61%) in the control municipalities. Allowing for death and drop-out, we analyzed data on participant mobility and mortality on the basis of 4034 individuals at baseline, 3690 at 18-month followup and 3383 at three-year follow-up (Table II). There was a significantly higher mortality rate in the intervention than in the control group for women (HR /1.461, p/0.009), whereas the opposite, Scand J Prim Health Care 2004; 22

although not significant (HR /0.877, p /0.279), was true for men. This analysis was adjusted for the covariates: age, mobility score at baseline and the 17 pairs of municipalities. When the analyses were stratified by sex, intervention had a significantly positive effect on women’s mobility (Fig. 1). If the dead were incorporated in the analyses almost identical associations were found (Table III). Intervention was effective with a number needed to treat of 19 for women in the 75-year age cohort and 18 in the 80-year age cohort. The intervention had no effect on the men’s mobility. Receiving home visits also had an effect on mobility, but again only among the women. Living alone had a negative effect on the men’s mobility. This was not seen for the women. Age and time had the expected declining effect on mobility, and there was no effect on mobility for the 17 pairs of municipalities.

DISCUSSION The main result of this study is that a brief, simple and feasible educational intervention towards professionals working with preventive home visits was associated with better functional ability among women. Furthermore, accepting and receiving preventive home visits was associated with better functional ability among women, but not among men. We targeted non-institutionalized individuals. The mortality rate over the three years was 12.9%, which is


109

Table II. Mortality and drop-out in study population during the three study years (1999 /2001). N /4034 Intervention 2,092 Participants

Men Baseline 1999 Survey 2000


Survey 2001/2 Total Women Baseline 1999 Survey 2000 Survey 2001/2 Total All Total mortality $

Drop-out

Died before drop-out

Control 1,942 Died after drop-out

Participants

933

Drop-out

Died before drop-out

Total Died after drop-out

852 31

57

2

88

845

1,785 27

64

8

77

761 4

755

1,606 8

676 33 (3.5)$

145 (15.5)$

1,431 35 (4.1)$

1,159

141 (16.5)$

1,090 38

48

6

74

1,073

2,249 39

40

6

46

1,011 8

993

2,084 8

959

1,952

44 (3.8)$

122 (10.5)$

45 (4.1)$

77 (3.7)$

267 (12.8)$ 12 279 (13.3)$

80 (4.1)$

86 (7.9)$ 227 (11.7)$ 16 243 (12.5)$

522 (12.9)$

( ) percentages.

similar to what is reported in other studies of in-home assessment of home-dwelling populations (4,21). The refusal of 30% of the eligible study participants may represent a weakness. However, preliminary analysis of the non-participants revealed no major differences in mortality between intervention and control municipalities at follow-up. A matched randomisation design was chosen to allow for the considerable variations in management and organisation. The results did not change when adjusted for municipality variation. The municipalities could not be blinded to the intervention, but data collection varied in both directions supporting that no systematic over-reporting from intervention municipalities took place. All participants knew that their municipality was taking part in a project, but they did not know whether they belonged to an intervention or control municipality. We observed no overall differences in response rates between intervention and control municipalities (10), which supports that most participants were blinded to the intervention. It was, however, impossible to avoid communication between intervention and control municipalities, even if no educational intervention took place in the control municipalities. During the study period some home visitors from both intervention and control municipalities exchanged experience at county meetings (not a part of this study). This could dilute some

of the intervention, but all these ‘control interventions’ would tend to underestimate positive effects. A strength of this study was that no major home visitor staff differences existed between intervention and control municipalities at baseline. Other study strengths include the high number of participants, the low drop-out, and the feasibility of the intervention owing to the structured guidelines that were easily implemented in regional education. A low drop-out rate was achieved through vigorous follow up and high motivation among all participating communities. A further study strength is the incorporation of a detailed cost-effectiveness analysis, the promising results of which will be published elsewhere. Women had a significantly higher mortality rate in the intervention than in the control municipalities. Intervention participants were not more disabled at baseline (10). As several studies report that preventive home visits are associated with lowered mortality rates, it is unlikely that our intervention is associated with a higher mortality as a side effect. Since death is associated with functional decline we considered the possibility of a survivor selection phenomenon, but discarded it for two reasons: Firstly, sensitivity analysis confirmed (data not shown) that the mortality and effect rates were far too small to generate a noticeable survivor selection effect. Secondly, such an effect would be in contrast to the fact that the effects are quite similar irrespective of how the Scand J Prim Health Care 2004; 22


110

M. Vass et al.

Fig. 1. Odds Ratios for having better functional ability after 3 years (N /4,034). Adjusted analyses including drop-outs without the dead (95% Confidence Intervals).

dead were included in the analysis. These conclusions are made possible by the recent methodological contribution of Dufouil et al. (20). The mortality was higher among drop-outs than among participants. It was not possible to incorporate this parameter in the statistical analyses, but there was no major difference in drop-out mortalities between the intervention and the control group. This study introduces a gender perspective in the inhome assessment discussion. A Danish study from the 1970s showed an insignificant effect on nursing home

relocations among women only (22). Studies have confirmed that women are more aware of local activities and more likely to adapt to lifestyle advice than men (5,23). Women may, moreover, in general be more conscious of healthy behavior throughout life because they have closer and more continuous contact with preventive and health services (24 /26). Almost all preventive home workers are women, which may also lead to gender differences in the impact of preventive messages. We do not know whether in-home assessment performed by men would

Table III. Effects on functional ability 1999 /2001. Men OR* Intervention vs. control Preventive home visits ( ]/1 vs. 0) Living alone vs. with others Time of study (baseline vs. 3 years) Age (younger vs. older)

CI95

SE of OR

Adjusted analyses including drop-outs 0.106 1.040 0.851 /1.270 0.984 0.802 /1.207 0.103 0.088 0.777 0.622 /0.970 0.096 1.345 1.170 /1.546 0.186 1.703 1.355 /2.109

Women p-value

Scand J Prim Health Care 2004; 22

SE of OR

p-value

1.076 /1.469 1.160 /1.595 0.853 /1.178 1.441 /1.763 1.689 /2.346

0.010 0.111 0.082 0.082 0.167

0.0040 0.0002 0.9765 B/.0001 B/.0001

the dead 1.265 1.081 /1.481 1.362 1.160 /1.560 0.999 0.849 /1.175 1.736 1.572 /1.917 2.022 1.713 /2.387

0.102 0.112 0.083 0.088 0.171

0.0034 0.0002 0.9883 B/.0001 B/.0001

without the dead 0.7031 1.257 0.8778 1.361 0.0309 1.002 B/.0001 1.594 B/.0001 1.990

Adjusted analyses including drop-outs and including Intervention vs. control 1.040 0.847 /1.277 0.109 0.7109 0.986 0.799 /1.212 0.105 0.8924 Preventive home visits ( ]/1 vs. 0) 0.091 0.0415 Living alone vs. with others 0.782 0.622 /0.983 0.107 B/.0001 Time of study (baseline vs. 3 years) 1.544 1.348 /1.769 0.197 B/.0001 Age (younger vs. older) 1.754 1.408 /2.185 * OR /0 is associated with better functional ability on the Mob-H scale.

OR

CI95



have a different outcome, and men may benefit more of other preventive strategies.

ACKNOWLEDGEMENTS This study was supported by grants from the Danish Medical Research Council, the Research Foundation for General Practice and Primary Care, Eastern Danish Research Forum, the County Value-Added Tax Foundation and the Danish Ministry of Social Affairs. We thank all participating municipalities and Eva Jepsen, Lisbeth Willemoes Sørensen, Annette Johannesen for following up questionnaires. We are indebted to Christian Cato Holm for data management and development of the municipality registration software.

REFERENCES 1. Stuck AE, Siu AL, Wieland GD, Adams J, Rubenstein LZ. Comprehensive geriatric assessment: a meta-analysis of controlled trials. Lancet 1993;342:1032 /6. 2. van Haastregt JC, Diederiks JP, van Rossum E, de Witte LP, Crebolder HF. Effects of preventive home visits to elderly people living in the community: systematic review. BMJ 2000;320:754 /8. 3. Elkan R, Kendrick D, Dewey M, et al. Effectiveness of home based support for older people: systematic review and meta-analysis. BMJ 2001;323:1 /8. 4. Stuck AE, Egger M, Hammer A, Minder CE, Beck JC. Home visits to prevent nursing home admission and functional decline in elderly people. Systematic review and meta-regression analysis. JAMA 2002;287:1022 /8. 5. Osler M, Jespersen NB. The effect of a community-based cardiovascular disease prevention project in a Danish municipality. Dan Med Bull 1993;40:485 /9. 6. Avlund K, Kreiner S, Schultz-Larsen K. Functional ability scales for the elderly. A validation study. Eur J Public Health 1996;6:35 /42. 7. Clark J. Preventive home visits to elderly people. Their effectiveness cannot be judged by randomised controlled trials. [Editorial]. BMJ 2001;323:708. 8. McNaugthon DB. A synthesis of qualitative home visiting research. Public Health Nurs 2000;17:405 /14. 9. Byles JE. A thorough going over: evidence for health assessments for older persons. Aust N Z J Public Health 2000;24:117 /23. 10. Vass M, Avlund K, Andersen CK, Hendriksen C, Keiding N. Preventive home visits to older people in Denmark. Aging Clin Exp Res 2002;14:509 /15.

111

11. Rubenstein LZ, Stuck AE. Preventive home visits for older people: defining criteria for success. Age Ageing 2001;30:107 /9. 12. Stuck AE, Walthert JM, Nikolaus T, Bu¨la CJ, Hohmann C, Beck JC. Risk factors for functional status decline in community-living elderly people: a systematic literature review. Soc Sci Med 1999;48:445 /69. 13. Avlund K, Damsgaard MT, Sakari-Rantala R, Laukkanen P, Schroll M. Tiredness in daily activities among nondisabled old people as determinant of onset of dependence of help. J Clin Epidemiol 2002;55:965 /73. 14. Avlund K, Damsgaard MT, Schroll M. Tiredness as determinant of subsequent use of social and health services. J Aging Health 2001;13:267 /86. 15. Podsiadlo D, Richardson S. The Timed ‘‘Up and Go’’: A test of basic functional mobility for frail elderly persons. J Am Geriatr Soc 1991;39:142 /8. 16. Williams IE, Junius U, Jones D, Fischer G, Sandholzer G, Vass M. An evidence-based approach to assessing older people in primary care. The Royal College of General Practitioners, Occasional Paper 82, 2002. 17. King AC, Rejeski J, Buchner D. Physical activity interventions targeting older adults. A critical review and recommendations. Am J Prev Med 1998;15:316 /33. 18. Modin S, Furhoff AK. Care by general practitioners and district nurses of patients receiving home nursing: a study from suburban Stockholm. Scand J Prim Health Care 2002;20:208 /12. 19. Avlund K, Thudium D, Davidsen M, Fuglsang-Sørensen B. Are self-ratings of functional ability reliable? Scand J Occup Ther 1995;2:10 /6. 20. Dufouil C, Clayton D, Spiegelhalter D. Analysis of longitudinal studies with death and dropout: a case study. I: General considerations and the marginal modelling approach. Stat Med. In press. 21. Egger M. When, where, and why do preventive home visits work? [Commentary]. BMJ 2001;323:8 /9. 22. Gunner-Svensson F, Ipsen J, Olsen J, Waldstrøm B. Prevention of relocation of the aged in nursing homes. Scand J Prim Health Care 1984;2:49 /56. 23. Schultz-Larsen K. Chronic disease and functional ability among 70-year old men and woman. A follow-up study of the 1914-population in Copenhagen County from 19641984 (In Danish). [Thesis]. Department of Social Medicine, University of Copenhagen; 1993. 24. Akiyama H, Elliott K, Antonucci TC. Same-sex and crosssex relationships. J Gerontol Psych Sci 1996;51B:P374 /82. 25. Shye D, Mullooly JP, Freeborn DK, Pope CR. Gender differences in the relationship between social network support and mortality: A longitudinal study of an elderly cohort. Soc Sci Med 1995;41:935 /47. 26. Burton LC, Shapiro S, German PS. Determinants of physical activity initiation and maintenance among community-dwelling older persons. Prev Med 1999;29:422 /30.

Scand J Prim Health Care 2004; 22