Systematic review of the links between human resource management practices and performance.

BACKGROUND: In recent years human resource management (HRM) has been seen as an important factor in the successful realisation of organisational change programmes. The UK NHS is undergoing substantial organisational change and there is a need to establish which human resource (HR) initiatives may be most effective.

OBJECTIVES: To assess the results from a wide-ranging series of systematic reviews of the evidence on HRM and performance. The first part assesses evidence on use of HRM in the UK and fidelity of practice implemented. The second part considers evidence for the impact of HRM practices on intermediate outcomes, which can impact on final outcomes, such as organisational performance or patient care.

DATA SOURCES: The following databases were searched: Applied Social Sciences Index and Abstracts (ASSIA), British Nursing Index (BNI), Business Source Premier, Campbell Collaboration, Cochrane Central Register of Controlled Trials (CENTRAL), Cochrane Database of Systematic Reviews (CDSR), Cumulative Index to Nursing and Allied Health Literature (CINAHL), Database of Abstracts of Reviews of Effectiveness (DARE), DH-Data, EMBASE, Health Management Information Consortium (HMIC), International Bibliography of the Social Sciences (IBSS), King's Fund database, MEDLINE, NHS Economic Evaluation Database (NHS EED), National Research Register (NRR), PREMEDLINE, PsycINFO, ReFeR, Social Sciences Citation Index (SSCI) and Science Citation Index (SCI). The searches were conducted in May/June 2006.

REVIEW METHODS: Broad categories of HRM interventions and intermediate outcomes were generated: 10 HRM categories and 12 intermediate outcome categories. Seven patient final outcomes were derived from the NHS Performance Indicators and the NHS Improvement Plan. The quality criteria used to select papers incorporated a longitudinal study design filter to provide evidence of the causal direction of relationships between HRM and relevant outcomes. Single HRM practices were considered. Within the health-specific literature, focus was on the impact of HRM on patient outcomes. Information is presented on the reliability of measures in each of the intermediate outcome areas.

RESULTS: Work design practices that enhance employee autonomy and control influenced a number of outcomes and there was consistent evidence for the positive impact of increased job control on employee outcomes, such as job satisfaction, absence and health. For employee participation, the small number of studies reviewed supported the involvement of employees in design/implementation of changes that affect their work. In health literature in particular, employee involvement through quality improvement teams resulted in improved patient outcomes. Findings were positive for the impact of training on the intended outcomes of the initiatives. Support for the impact of performance management practices was apparent, in particular feedback on performance outcomes and the use of participative goal setting. Strong associations were found among all intermediate outcomes, and the relationship between most intermediate behaviours and outcomes were significant.

LIMITATIONS: Limited evidence was available on the use of HRM and on the implementation of policy. Also, the specific practices studied within each HRM category differ so there was little evidence to show whether similar practices have the same effects in health and non-health settings.

CONCLUSIONS: Some potentially effective practices for both health and non-health areas were identified, and HRM methods could be used to support change processes within the NHS; the findings relating to work organisation are particularly promising with regard to changes in methods of service delivery. Using training to support the implementation of change is highlighted. However, future multilevel studies that embrace the individual, team and organisational level are needed. Studies should look into interventions aimed at improving HR outcomes and performance, and allow for pre- and post-intervention measurement of practices and outcomes.