Figure 1 describes the summary of the literature search results and article inclusion and exclusion at each level. Sixty-three studies met inclusion criteria. Based on the decision to focus on outcomes with at least 3 supporting studies, data from 37 studies (14 RCTs and 23 observational studies) were included in outcome aggregation. A summary of study design, study groups, study purpose, patient population, outcomes, and quality of individual studies are included in Supplementary Table 1 (available online at www.npjournal.org).
Eleven patient outcomes were identified, for which results were reported in at least 3 studies. Quality of care measures reported included patient satisfaction with provider/care, patient self-assessment of perceived health status, functional status, number of unexpected ED visits, hospitalization, duration of ventilation, and hospital LOS. Effectiveness of care was represented by patient blood pressure, blood glucose, and serum lipid levels. Mortality was the only safety outcome reported.
Most studies were conducted in urban rather than rural areas. Care delivery settings varied and included primary care offices and clinics, private homes, long-term care facilities, and inpatient acute and critical care areas. NPs were, at most, master's prepared, but years of professional experience were not reported for any providers. MDs working alone or in a group were the typical comparison group. A number of studies compared outcomes from teams that included NPs to those of teams inclusive of medical trainees (interns, residents, and fellows).[33–44] Since medical care provided by medical trainees is supervised by an attending MD who retains accountability for patient care, it was presumed that care provided by trainees reflected the influence of the attending MD.
Where not otherwise noted, it was presumed for studies conducted in inpatient hospital settings that NPs and MDs consulted daily. This frequency of consultation is common in that setting. However, in 5 of the RCTs and 5 of the observational studies, it appeared that NPs provided care with very little or no MD consultation.[45–54]
When comparing outcomes for quality of care provided by NPs with care involving only MDs, the strength of evidence was high, indicating similar patient satisfaction with provider/care,[33,46,48,54–56] self-report of perceived health status,[34,41,47,48,50,55,57] functional status,[34,50,57–64] numbers of unexpected ED visits,[47,49,51,53,57] and hospitalization rates.[36,37,40,44,47,51–53,57,61,64] A moderate strength of evidence indicated that care involving NPs was similar to care involving only MDs in terms of hospital LOS.[33–40,42–44,51,53,65–67] And a low strength of evidence indicated that duration of ventilation (for adults) was similar for care involving NPs compared with care involving only MDs.[35,38,43]
When comparing outcomes related to effectiveness of care by NPs with care involving only MDs, the strength of evidence was high, indicating similar patient outcomes for blood glucose[45,46,47,48,55] and blood pressure.[45,46,48,55] There was high strength of evidence of better effectiveness of care on the outcome of patient serum lipids from care provided by NPs than from care involving only MDs.[45,55,69]
A detailed summary of the aggregated outcomes and evidence for NPs can be found online.
Journal for Nurse Practitioners. 2013;9(8):492-500.e13. © 2013 Elsevier Science, Inc.