Important

Return to table of contentscare needs fall within the nurse practitioner’s scope ofpractice. In this role, nurse practitioners provide importantservices and free physicians’ time to focus on thehigh-level diagnostic and therapeutic services for whichthey have been trained. Consultation and collaborationare essential skills for nurse practitioners, as they are forall members of the health care team.The Board of Directors of the National Organizationof Nurse Practitioner FacultiesWashington, DCIn Reply: Many of the above comments challenge thegeneralizability of our findings. We appreciate this limitationbut believe that our results are suggestive enoughto open more active discussion about the appropriaterole of various primary care practitioners. Dr Bagley,Dr Chan-Tack, Dr Hicks, and Dr Rayburn all raise questionsabout the supposed healthy nature of our populationand the relative ease of using practice guidelinesto care for patients with chronic conditions. The highburden of illness in the population is reflected in theirSF-36 scores, which were 35% lower on average than anational sample of similar age and sex. Moreover, veryfew patients had only a single previously diagnosed condition.Less than 5% (58 of 1316) of the study patientswere treated in the first 6 months of the study for only1 of the chronic conditions (or for a related diagnosis)or had a general medical examination. Hicks andDr Nasir question the poor diabetic control both typesof practitioners achieved. The good results achievedwith a Hispanic population in Colorado are commendable.We do not know how ours population differs fromthat in Colorado, but we do know ours was poor, minority,exceptionally transient, and had a high burden ofillness. The number of patients scheduled and seen wassimilar in both physician and nurse practitioner clinicsites; the higher enrollment in the nurse practitionercohort attests to the greater number of appointmentsavailable in the newer nurse practitioner clinic. Bagleyquestions the validity of the SF-36 in detecting differencesover time in patients. This instrument has beenused in many other studies to detect such change. TheSF-36 was certainly sensitive enough to track improvementin the study patients’ conditions from the timeof their initial emergency department visit through theseveral follow-up points. Moreover, we also used physiologicalmeasures and utilization data to complete ouranalysis. We agree that the brief time frame for thestudy was a problem. It represented a trade-off betweensample loss and time to observe an effect. All practitionersin the study were salaried, part-time employeesof hospital-based primary care clinics, and all were fulltimefaculty in either the medical or nursing school. Nohouse staff were involved. Rayburn raises questions ofstatistical power. Studies designed to show equivalencyrather than differences require careful attention to thisissue, which we addressed at some length. In planningthe study we anticipated the issue of statistical powerin calculating our sample size because the primaryhypothesis was for no difference between physician andnurse practitioner practices. We confirmed our originalcalculations with the actual data and concludedthat the findings would not have changed with a largersample. Indeed, it would require a very large sample toproduce statistically significant differences, and thesedifferences would not consistently favor the physiciangroup. Intention-to-treat analysis was not used becauseit would exacerbate the problems of a study designedto assess comparability; it fosters a conservative test ofdifferences. Baseline physiological measurements wouldhave been very helpful, but the logistics of randomizationprevented us from obtaining them. No single studywill satisfactorily resolve a controversial issue. We hopethis study will encourage similar efforts to test new primarycare models.Mary O. Mundinger, DrPHColumbia University School of Nursing, New York, NYRobert L. Kane, MDUniversity of Minnesota School of Public Health,Minneapolis, MNIn Reply: I agree with most of Dr Poplin’s comments.I disagree with her contention that we should encouragenurse practitioners to practice independently, withoutrequiring them to prove that they measure up tophysicians in caring for very sick patients. Our societyrequires the proponents of new drugs and new tests toprove that the new technology is as effective as theestablished technology. In this way, those who pay forhealth care, or those who use it, can decide whetherthe new technology is effective in a specific situation.Shouldn’t patients know if physicians are better thannurse practitioners at some aspect of primary care sothat they can decide when to ask for a consultation?The letter from the Board of Directors of the NationalOrganization of Nurse Practitioner Faculties takes issuewith several examples that I used to support my critiqueof the study by Dr Mundinger and colleagues. I contendthat utilization of health care after 1 year is a measureof the quantity of health care, not the quality of itsoutcomes. Although Hispanic populations may havea higher prevalence of some diseases, they have lowermortality rates than their socioeconomic status wouldpredict. The Board of Directors of the National Orga-Scope of Practice Data Series: Nurse practitioners • Appendix72