How Do Nurse Practitioners Compare to MDs as Primary Care Providers? Rather Well
From 1998 to 1999, researchers at the Columbia University School of Nursing conducted the second phase of a study comparing nurse practitioners and physicians as primary care providers.
During Phase I, researchers assigned patients randomly to either a nurse practitioner or a physician at Columbia Presbyterian Medical Center. Researchers had found no significant differences in the health status or health services utilization of patients in the two study groups, after one year.
To see if the findings were maintained over an additional year, in Phase II of the study researchers collected additional data through 756 patient interviews (439 with the nurse practitioner group and 318 with the physician group).
- Patients who were assigned to nurse practitioners were similar demographically to patients assigned to physicians.
- In the year before this data collection:
- Thirty-three percent of patients received care only at the assigned clinic.
- Six percent received care at the assigned clinic and another provider.
- Twenty-seven percent only sought care elsewhere.
- Thirty-two percent did not seek primary health care at all.
- Researchers concurred with the preliminary results: in an ambulatory care situation where nurse practitioners have the same authority, responsibility, productivity and administrative requirements as physicians, patient outcomes are comparable.
The Robert Wood Johnson Foundation (RWJF) supported this project through a grant of $162,075.
Assuring adequate access to health care for all segments of society continues to be a major challenge for policymakers and the health care industry. NPs serving as primary care providers have the potential to improve access to care, but there was a lack of definitive data on the impact of NPs in this role.
Columbia University School of Nursing assumed a leadership role in advanced nursing practice in 1994 with the opening of the Center for Advanced Practice, a nurse practitioner primary care clinic, one of six community-based primary care practices maintained by Columbia Presbyterian Medical Center.
It serves a primarily Hispanic Medicaid population of more than 3,200 adult and pediatric patients in the Washington Heights-Inwood section of Manhattan. It is staffed by faculty nurse practitioners who have been granted hospital-admitting privileges by the Medical Board of the medical center.
During Phase I of this project, the Columbia University School of Nursing conducted a randomized study to compare NPs and MDs as primary care providers. This study, conducted from August 1995 to October 1997, was entitled "The Evaluation of Primary Care in Washington Heights." It was funded by the New York State Department of Health, the Division of Nursing of the US Department of Health and Human Services, and the Leslie Samuels and Fan Fox Foundation.
The study enrolled 1,316 patients who were referred from two emergency departments and one urgent care center at Columbia Presbyterian Medical Center. These patients were assigned randomly to either a nurse practitioner (806) or a physician (510) at Columbia Presbyterian Medical Center.
Baseline data included the patients' functional health status and whether they had one of three chronic conditions targeted for study: asthma, diabetes, or hypertension. Patients were interviewed six months after their initial appointment and health services utilization data were recorded at six months and one year after the initial appointment.
As reported in The Journal of the American Medical Association (JAMA), no significant differences were found in the health status of patients who saw nurse practitioners and those who saw physicians. No significant differences were found in health services utilization after six months or one year. Following the initial appointment, there were no differences in patient satisfaction with primary care. At six months, satisfaction ratings differed only for provider attributes, with physicians rated slightly higher (4.2 vs. 4.1 for NPs on a scale where 5 equals excellent).
The researchers concluded that patient outcomes were comparable in an ambulatory care situation in which patients were randomly assigned to either NPs or MDs, and where NPs had the same authority, responsibilities, productivity, administrative requirements, and patient population as MDs.
RWJF also helped supported an evaluation of the Columbia Advanced Practice Nurse Associates (CAPNA) practice in midtown Manhattan. See Grant Results on ID# 032045.
To determine whether earlier findings would be maintained, Phase II built on the original project by collecting two-year follow-up data. The objectives were: (1) to compare two-year outcomes for patients retained by nurse practitioners in primary care practice with those of patients retained in physician primary care practices; and (2) to compare outcomes for patients who received the assigned treatment from either an MD or NP practice over a two-year period with those who did not receive it.
In essence, the second objective shifted from a comparison of MD and NP treatment to considering who is most likely to remain with a given primary care provider long enough to receive continuity of primary care, and ultimately to the outcome of receiving, or not receiving, that care.
The researchers conducted 756 interviews (439 NP patients and 318 MD patients) between April 1998 and December 1999, representing a 66.3 percent response rate (65.9 percent of the original NP patients and 66.9 percent of the original MD patients). These interviews included a standard questionnaire on health status and, for those with a diagnosis of asthma, diabetes, or hypertension, a condition-specific questionnaire; and questionnaires on satisfaction and on health services utilization. The interviews were conducted by bilingual lay interviewers from the community.
The researchers validated the health services utilization data obtained from patient surveys with hospital data and six months of Medicaid data from before and after the baseline measure. The researchers were unable to conduct the final data analysis during the grant period because collecting data from the transient population studied took longer than expected.
This phase of the project was also supported by the United Hospital Fund ($60,000).
Findings for the 756 patients indicate that:
- Patients who were assigned to nurse practitioners were similar to those who were assigned to physicians. As with the 1,316 patients originally enrolled in the study, there were no statistically significant differences in the demographic characteristics among the two groups in terms of age, sex, race, and ethnicity, whether they spoke Spanish, or the presence of three medical conditions studied: asthma, diabetes, and hypertension.
- According to self-reported data on the year prior to data collection, 33.3 percent of patients received care only at the assigned clinic; 6.3 percent received care at the assigned clinic and through another provider; 27.1 percent only sought care elsewhere; and 32.4 percent did not seek primary health care at all. The main reasons for not seeking primary health care were "not sick" (53.6 percent MD patients, 52 percent NP patients), and "no insurance/money" (18.6 percent MD patients, 17.6 percent NP patients). Fewer NP patients than MD patients returned to the assigned clinic, possibly because the NP practice relocated during the study period.
After the grant period ended, the researchers completed data analysis for two sub-samples of patients: (1) 217 patients who received primary care only from the practices to which they were randomized between 12 and 24 months after their initial appointment (99 NP patients and 118 MD patients); and (2) 406 patients who saw their assigned provider at least once in the two years after the initial visit and received primary care only from that provider (222 NP patients and 184 MD patients).
- Analysis of data for the first sub-sample of patients, showed that:
- At the start of the study, the NP and MD patients did not differ in demographics or health status.
- At two years after the initial visit, the two groups did not differ in any outcome measured:
- NP and MD patients with hypertension, diabetes, and asthma did not differ in their two-year blood pressure, blood sugar, or peak flow values (maximum air flow from the lungs).
- At the start of the study, there were no differences between the NP and MD groups.
- At two years after the initial visit, the only outcome on which the two groups differed was average number of primary care visits during year two. NP patients averaged 1.76 visits and MD patients averaged 2.5 visits.
Due to the small sample size, the researchers had a limited ability to detect statistically significant differences between patients randomized to the nurse practitioner and physician practices.
The researchers presented these results at the annual meeting of the Western Institute of Nursing in April 2001. They are working on a manuscript. Columbia School of Nursing covered the cost of staff time for conducting analyses and preparing publications and presentations.
- Follow-up on transient populations is difficult; studies should be designed to maximize sample size; additional funding may be needed for studies of these populations. These study subjects were difficult to follow-up because they moved frequently, often did not have phones, and frequently left the area to visit their native countries. The researchers originally planned to conduct the two-year survey only on patients who had completed six-month interviews, but ultimately decided to try to interview all patients in order to maximize the sample size. Doing so boosted the response rate from 60.3 percent to 66.3 percent. Even so, data collection difficulties kept investigators from finishing data analysis with grant funds during the grant period.
- Utilization and satisfaction instruments should be geared for the study population subjects. Subjects in this trial had difficulty understanding the health care system; when the nurse practitioner practice was re-located, some patients assumed they were using a different practice though they were using the same practice in its new location. Also, more intensive and frequent follow-up, which is expensive, is required.
- When studying populations for whom English is not the primary language, use bilingual interviewers who are part of the community, and be prepared to train them. Bilingual interviewers greatly facilitate access to patients; however, individuals selected from a local community are unlikely to know much about conducting research, and must be thoroughly trained.
- Where possible, secure objective data sources to validate self-reported data. Because data on health services utilization reported by patients themselves can be subject to recall errors, the researchers used Columbia Presbyterian Medical Center data to validate patient self-report data. This increased the usefulness of the patient surveys. The researchers found that patients seeing physicians over-reported returning to their assigned provider while nurse practitioner patients under-reported it.
AFTER THE GRANT
The researchers have obtained Medicaid data for the entire two years of the study from the state and they hope to analyze it to help further validate health-services utilization data.
GRANT DETAILS & CONTACT INFORMATION
Bridge Funding for the Evaluation of a Primary Care Practice in Washington Heights
Columbia University School of Nursing (New York, NY)
Dates: June 1998 to December 1999
Mary O. Mundinger, R.N., Dr.P.H.
(Current as of date of this report; as provided by grantee organization; not verified by RWJF; items not available from RWJF.)
Presentations and Testimony
Mary O. Mundinger, Elizabeth R. Lenz, Sarah Hopkins, and Susan Lin, "Comparing NP and MD Primary Care Outcomes: Two-Year Follow-Up," at the Western Institute of Nursing Annual Meeting, April 21, 2001, Seattle, Wash.
Mary O. Mundinger, Elizabeth R. Lenz, Sarah Hopkins, and Susan Lin, "Comparing NP and MD Primary Care Outcomes: Two-Year Follow-Up," at the Academy for Health Services Research Annual Research Conference, June 12, 2001, Atlanta, Ga.
Report prepared by: Lori De Milto
Reviewed by: Robert Crum
Reviewed by: Karyn Feiden
Reviewed by: Molly McKaughan
Program Officer: Susan Hassmiller