Effects of psychosocial interventions on quality of life in adult cancer patients: meta analysis of 37 published controlled outcome studies

Abstract

The present meta analysis summarized the results of 37 published, controlled studies that investigated the effectiveness of psychosocial interventions on quality of life (QoL) in adult cancer patients. The overall effect size of psychosocial interventions and the effect of potential moderating variables such as type and duration of intervention, sociodemographic and clinical parameters, characteristics of QoL measurement, and methodological quality of the selected studies were calculated using a meta analysis model suggested by Hunter and Schmidt. The overall effect size was 0.31 (correlation equivalent delta) which corresponds to a standardized mean difference of d=0.65 (N=3120 cancer patients). The most important moderating variable was duration of psychosocial intervention with durations of more than 12 weeks being significantly more effective than interventions of shorter duration. The meta analytical findings support the usefulness of psychosocial interventions for improving QoL in adult cancer patients.

Introduction

There is considerable evidence suggesting that cancer patients suffer from substantial and long-term psychological distress associated with different forms of cancer and its medical treatment [1]. The psychosocial management of adjustment problems experienced by people with cancer seems to be an obvious requirement for a more effective treatment of the disease. Nevertheless, there is an ongoing debate about whether and to what extent psychooncological care can be effective in patients suffering from cancer. These questions were comprehensively addressed by narrative reviews in several previous investigations [2], [3], [4], [5], [6], [7], [8], [9].

However, there are only a few studies that examined the overall effectiveness of psychosocial interventions quantitatively by means of meta analysis [10], [11], [12]. The small number of quantitative meta analyses might be due to the fact that the single intervention studies are difficult to compare. Differences in clinical characteristics and methodological design can easily obscure the investigated effects of psychosocial interventions on outcome measures. Potential confounding variables are for example, (1) the patients’ clinical and demographic characteristics such as type of diagnosis, stage and course of the disease, medical treatment, age, gender, and educational level; (2) type and duration of psychosocial interventions; (3) selection and operationalization of outcome measures used to assess the effectiveness of psychosocial interventions; and (4) the methodological quality of intervention studies indicated by the choice of control groups, randomization status of treatment conditions, or documentation of experimental and statistical designs and procedures. Meta analyses can quantify the impact of such possible moderating variables in terms of correlational relationships, although causal explanations of these relationships cannot be provided.

The present meta analysis tried to reduce the problem of different outcome measures by restricting relevant criteria to subjective quality of life (QoL) excluding other outcome criteria such as psychiatric syndromes (fear, depression), survival time, relapse rate or side effects of medical treatment. QoL was chosen because of the clinical relevance of this concept [13], its multidimensional structure reflecting the impact of cancer on several areas of the patients’ life [14], and its sensitivity to psychotherapeutically induced effects [2], [6], [9]. However, QoL is no well-defined homogeneous concept so that a possible significant impact of the different conceptualizations of QoL on outcome measurement should be controlled.

The present meta analysis sought to address the following issues: (1) What is the overall effect size of psychosocial interventions on QoL in adult cancer patients? (2) Do various types of psychosocial interventions create different effects? (3) Does effect size vary as a function of duration of psychosocial intervention? (4) Do sociodemographic and clinical parameters moderate these effects? (5) Do different conceptualizations of QoL reveal a significant impact on the effect size? (6) Does the methodological quality of intervention studies modify the effect size?