Reporting of critical information in studies of pharmacists in HIV care.

OBJECTIVE

To evaluate manuscripts documenting HIV pharmacist interventions and assess adequacy of reporting as defined by CONSORT and STROBE criteria.

METHODS

PubMed, EMBASE, Cochrane Library, Web of Science, BIOSIS Previews, and PsycINFO databases were searched from inception - 1 June 2011. Studies were included if pharmacists performed an intervention to improve HIV patient care, and the study evaluated the intervention's impact. Qualitative studies, non-English language reports, abstracts and studies where the pharmacist did not intervene were excluded. Manuscripts were independently evaluated by two reviewers for the presence, absence or lack of applicability of STROBE (observational studies) or CONSORT (randomized studies) criteria, for presence or absence of description of pharmacist's duties, CD4+ cell count, HIV viral load and adherence measurement. Reviewers met to discuss the rationale behind their evaluation; a third arbiter was consulted when reviewers could not agree on a particular criterion.

KEY FINDINGS

Twenty-two manuscripts met inclusion criteria. Observational studies of HIV pharmacists (n = 19) included 56% of applicable STROBE criteria. Randomized studies of HIV pharmacists (n = 3) adhered more closely to CONSORT reporting guidelines (average 80% of applicable criteria). Manuscripts published after 2004 more consistently evaluated pharmacist impact on HIV outcomes such as CD4+ and viral load.

CONCLUSIONS

Thorough reporting increases the reader's ability to critically evaluate manuscripts of HIV pharmacist services. Increasing pharmacist awareness of manuscript guidelines such as CONSORT and STROBE may improve clarity of reporting in studies of HIV pharmacist interventions and clinical programmes.