The influence of outpatient comprehensive geriatric assessment on survival: a meta-analysis

Abstract

Although outpatient Comprehensive Geriatric Assessment (CGA) has shown certain benefits in functional status and quality of life by many randomized controlled trials, no survival benefit has been reported. We hypothesized that the lack of survival benefit may be due to insufficient power of individual trials. In order to assess the influence of outpatient CGA on survival of older persons, we performed a meta-analysis of all randomized controlled trials of outpatient CGA. Nine studies consisting of 3750 subjects fulfilled the predetermined eligible criteria and were included in the meta-analysis. Combined mortality risk ratio with outpatient CGA intervention compared to usual care group was 0.95 (95% confidence interval, CI 0.82–1.12, P = 0.62). Treatment effects were homogeneous across the trials. This meta-analysis did not demonstrate survival benefit for outpatient CGA. Inadequate statistical power is unlikely to explain the results. Future researches of outpatient CGA should focus on coordinated and standardized measurement of outcomes related to functional status, institutionalization rate, and quality of life.

Introduction

The Comprehensive Geriatric Assessment (CGA) emerged during the1980s as an important strategy to improve care for elders with complex medical, psychosocial, and functional problems (Epstein et al., 1987). CGA is more effective when performed on specialized hospital units than in other settings. The more intensive the CGA intervention and the more control the intervening team has over patient care, the more likely CGA will have a beneficial effect on outcomes (Applegate and Burns, 1996). Focus of recent investigations has been shifted to outpatient CGA due to high costs of inpatient care (Boult et al., 2001). However, the studies of outpatient CGA are impeded by low statistical power, non-uniform eligibility or targeting of patients, and heterogeneous outcomes measures (Applegate et al., 1991, Boult et al., 2001), its benefits are not consistently demonstrated.

Survival is one of the most commonly reported outcomes in clinical studies. Stuck et al. (1993) reported in a meta-analysis that outpatient CGA did not improve survival compared to usual care despite significant survival benefit was observed in inpatient and home-based CGA. However, two of the four trials of outpatient CGA included in this meta-analysis were criticized because the subjects were relatively healthy and not targeted as being at high risk for decline. Since the publication of that meta-analysis, additional studies of outpatient CGA were conducted with greater attention to targeting frail subjects. These studies showed that the intervention improved various outcomes in older persons including mental health (Rubin et al., 1993, Burns et al., 1995, Burns et al., 2000, Reuben et al., 1999, Boult et al., 2001, Cohen et al., 2002) and functional status (Rubin et al., 1993, Reuben et al., 1999, Boult et al., 2001) at no increased cost (Rubin et al., 1993, Toseland et al., 1997, Boult et al., 2001, Cohen et al., 2002). In addition, more than half of these trials about outpatient CGA demonstrated a trend to improve survival (Epstein et al., 1990, Silverman et al., 1995, Toseland et al., 1997, Reuben et al., 1999, Burns et al., 2000, Boult et al., 2001) although no one showed a significant survival benefit.

We hypothesized that the lack of demonstrated association between outpatient CGA and reduced mortality may be due to the sample size limitations of these individual trials. To test this hypothesis, we conducted a meta-analysis of all randomized, controlled trials of outpatient CGA, to determine whether the collective statistical power of these studies would demonstrate survival benefit for older patients.