This review found that general practice-based health checks were associated with statistically significant but clinically small improvements in surrogate outcomes such as serum cholesterol and blood pressure. The authors' conclusions appropriately reflect the evidence presented. Limited reporting and use of surrogate outcomes make it difficult to interpret the relevance of the findings for general population health checks.

To assess the effects of health checks in general practice settings.

PubMed, EMBASE and Cochrane Central Register of Controlled Trials were searched to October 2012. Citation searches were performed using SCOPUS and Google Scholar. Search strategies were reported to be available on request. Only articles in English were included.

Randomised controlled trials (RCTs) or pseudo-randomised trials of general health checks (with or without subsequent intervention) versus no health check were eligible for inclusion. Trials had to involve a middle-aged population (mean age 35 to 65 years) and be performed in general practice settings. Outcomes of interest were total mortality, cardiovascular disease (CVD) mortality, blood pressure, total cholesterol, body mass index (BMI) and smoking status.

In the included trials, health checks were performed by practice staff or trained personnel. Four trials targeted general populations and two recruited high-risk patients and their partners. Various interventions were implemented after the health check. Further details of studies were available on request.

The authors did not state how many reviewers selected studies for inclusion.

Study quality was assessed using the Cochrane risk of bias tool. The authors did not state how many reviewers were involved.

Data were extracted to calculate mean differences (MD) for continuous outcomes and odds ratios (OR) for dichotomous outcomes, both with associated 95% confidence intervals (CI). For the one cluster RCT included, only data that had been adjusted for clustering effects were used. It appeared that data were extracted by one reviewer and independently checked by two others.

Data were pooled by meta-analysis using a random-effects model. Statistical heterogeneity was assessed using Χ² tests and the Ι² statistic. Substantial heterogeneity was defined as Ι²>85%. Prespecified subgroup analyses (based on different recruitment strategies and length of follow-up) were used to explore heterogeneity. Publication bias was assessed using funnel plots.

To assess the effect of study setting, non-practice-based studies were identified from another systematic review (see Other Publications of Related Interest). Data were extracted from these studies and random-effects meta-regression was used to test whether health check effects differed between settings.

Six RCTs were included (one was a cluster RCT). The number of participants was not reported. Follow-up ranged from one to 10 years. Risk of bias varied between trials. The main limitations were lack of blinding and loss to follow-up.

Health checks were associated with small but statistically significant improvements in total cholesterol (MD -0.13 mmol/L, 95% CI -0.19 to -0.07; four trials; Ι²=52%), systolic blood pressure (MD -3.65 mmHg, 95% CI -6.5 to -0.81; four trials; Ι²=94%), diastolic blood pressure (MD -1.79 mmHg, 95% CI -2.93 to -0.64; four trials;, Ι²=88%) and BMI (MD -0.45 kg/m², 95% CI -0.66 to -0.24; three trials; Ι²=0%). Odds of a patient remaining at high risk (as defined in the trial) at the end of the study period were significantly reduced by health checks for cholesterol, blood pressure and BMI but not for smoking.

There was no significant difference between health check and control groups for total mortality (OR 1.03, 95% CI 0.9 to 1.18; four trials; Ι²=7%). CVD mortality was significantly higher in the health check group (OR 1.30, 95% CI 1.02 to 1.66; three trials; Ι²=7%).

Heterogeneity was variable and partly explained by differences in recruitment and follow-up (details in the paper). There was no evidence of publication bias. Results from non-practice-based trials were presented. Meta-regression found no significant differences in effects between practice-based and non-practice-based trials.

General practice-based health checks were associated with statistically significant but clinically small improvements in surrogate outcomes, especially for high-risk patients.

The review question and inclusion criteria were clear. The search covered a range of relevant sources. The restriction by language was a limitation. There was no apparent attempt to locate unpublished studies but no evidence of publication bias was found. Measures to minimise reviewer errors or bias were not reported except for data extraction. Participant and intervention details were not reported.

Quality of included trials was assessed using appropriate criteria and the results were used in the synthesis. The meta-analysis used standard methods and included an exploration of heterogeneity and a comparison of trials in different settings. The authors discussed limitations of the evidence, including that trials assessed surrogate outcomes and were not designed to assess effects on mortality.

The authors' conclusions appropriately reflect the evidence presented. The included trials involved a mixture of high-risk and general population groups. Limited reporting and use of surrogate outcomes make it difficult to interpret the relevance of the findings for general population health checks.

The authors did not state any specific implications for practice or further research beyond saying that more evidence was required on the effectiveness of health checks.

First author funded by the China Scholarship Council.

Krogsboll LT, Jorgensen KJ, Gronhoj Larsen C, Gotzsche PC. General health checks in adults for reducing morbidity and mortality from disease: Cochrane systematic review and meta-analysis. BMJ 2012; 345: 7191.