A total of 99 studies were included in the review. Some studies evaluated more than one type of human resource management intervention. Most studies were prospective cohorts with historical controls or retrospective cohorts; duration ranged from six weeks to seven years.
Work design (14 studies): Intervention content varied among included studies. Outcomes management and skill enhancement interventions had inconsistent results, but other job and work design interventions had positive effects on staff satisfaction, patient satisfaction and waiting times.
Staffing (63 studies): Evidence on the impact of a reduced or lower number of working hours suggested a possible negative effect on patient satisfaction, but not on other outcomes such as mortality, length of stay, medication error or adverse events and other outcomes (19 studies). The evidence on staff to patient ratios was mixed. Most interventions evaluating the introduction of new specially trained physicians or new nurse services reported positive outcomes, but results were less consistent for the introduction of nurse practitioners to teams with physicians (13 studies). No adverse effects were found relating to the replacement of one type of physician for another or for replacing physicians with senior nursing staff; staffing by registered nurses rather than replacement by other grades had a positive effect on mortality, length of stay and some infection rates but limited impact on other outcomes (23 studies). The evidence on the effects of transfer or relocation of staff or services was small and highly heterogeneous (eight studies).
Training and development (16 studies): Six of seven studies of training interventions that focused on infection control reported significant reductions in infection rates. Several studies also reported that training had a positive effect on staff knowledge and mortality rates.
Compensation and rewards (three studies): The small numbers of interventions were very diverse and reported mixed outcomes, limiting the possibility of drawing a conclusion about any causal relationship between compensation/rewards and patient outcomes.
Employee involvement (five studies): The included studies reported consistently positive results for the impact of 'quality circles' on a range of patient outcomes.
Performance management and appraisal (seven studies): The included studies reported consistently positive results for the impact of feedback (alone or with education/instruction) on infection rates and patient satisfaction.