Effective nurse-patient relationships in mental health care: A systematic review of interventions to improve the therapeutic alliance.

Only eight studies were included, so the review cannot identify which intervention works best. Six of the eight studies were rated weak methodologically, with limited randomisation, blinding, confounder control and reporting of withdrawals. Four studies found no significant difference in alliance, and positive findings came from weak or moderate-quality designs. Most studies measured alliance with the Working Alliance Inventory, a therapy tool that may not fit nurse-patient key-worker relationships. Interventions were often group-based, while alliance is a dyadic relationship between one nurse and one patient. Few studies checked fidelity to the intervention, so null results may reflect poor implementation rather than ineffective content.

Do not conclude that psychological formulation, reflective groups or shared activity improve the nurse-patient alliance because some studies reported gains. The review found no intervention confirmed effective, and most studies were weak; positive findings came from small, under-controlled designs.