INTRODUCTION

As health care reform progresses, nurse practitioners (NP) are increasingly scrutinized for the care they deliver. This scrutiny comes from both the medical community and from insurance and managed care organizations interested in quality and quantity of services delivered to their customers. To date, however, these same insurance companies and managed care organizations have, in reality, done very little to monitor the quality and quantity of care delivered by physicians much less NPs. However, with greater emphasis on the provision of primary care by NPs these same companies will monitor the efficiency and effectiveness of NPs. Utilization of resources through referrals by NPs may be targeted as a means for cost savings compared to referral rates of physicians. Since referral patterns are one aspect of cost effectiveness of a provider, it then becomes important to measure these patterns in order to measure economic outcomes of care.

Is it really more cost effective for NPs to treat a cohort of patients if referral and diagnostic patterns are greater than a physician? If not, then it makes little economic sense for these companies to reimburse the NP as a primary care provider (PCP) at the same rate as physicians.

LITERATURE REVIEW

Several studies have documented the cost effectiveness and quality of care delivered by NPs. The Office of Technology Assessment (₁), the American Nurses’ Association metanalysis (₂), other studies (₃,₄,₅,₆) have shown that NPs provide equal to or better care than that delivered by physicians practicing alone. However, there are little data regarding referral patterns of NPs. Nurse practitioner services are consistently obscured by data collection methods that focus on the collaborating physician. NP productivity, prescriptive practice, and revenue generation are recorded under a physician’s name and provider number, both at the insurance level and the pharmacy level (₇). Nevertheless, some studies have frivolously attempted to define what NPs do and how they do it.

Record and Schweitzer (₈) reported that NPs can manage 80% of adult and 90% of pediatric primary care currently delivered by physicians. A most recent study by Nurse Practitioners Education Associates (7) surveyed 776 NPs from across the United States regarding their practices. Results in this study showed that 17% of the nurse practitioners studied did not consult with a physician on any patient, 47% consult on 10-20% of patient encounters, 25% consult on 5-10%, and 11% consult on less than 5% of all encounters. In this particular sample 70% were practicing in an ambulatory setting and 78.7% were either family or adult nurse practitioners. Though this study seems to be a retrospective study based on NP estimations of referral patterns it provides insight into these referral/consultation patterns.

Hooker and McCaig (₉) looked at the extent to which physician assistants and/or NPs were a source of health care delivery in emergency departments in the United States. They indicated that physician assistants and NPs are involved in care for almost 4% of all emergency visits nationally and manage a wider range of conditions than has been previously reported. To their dismay, they were unable to locate any study that identified the range of emergency services that could be provided by NPs.

PROJECT PURPOSE

The purpose of this study was to describe NP referral patterns in two primary care practices which also provide occupational health care. Specifically, this study was intended to discover reasons why NPs refer patients; the types of specialists to whom they most frequently refer; the types of diagnostic procedures requested; and the percentage of patients referred to specialists.

SETTING

This study was conducted at two academic nurse practitioner managed clinics in large metropolitan areas. Both clinics are directed by doctorally prepared NPs who are professors of clinical nursing in their respective schools.

University of Tennessee-Memphis

University of Texas Health Services

PROCEDURE

Data for this study were simultaneously collected in the two sites during the months of January through March, 1997. All patients who received referrals by NP’s during these months were counted in this quality assessment project. A form was developed by the two sites after an initial discussion of content and a subsequent pilot study (Figure A).

Figure 1

Figure A: Patient Referral Data Collection

The NP completed the form at the time of visit for every patient who was referred to an internal (i.e. in-house physician support) or external provider (i.e. physician specialist, counseling) or service (i.e. radiology or imaging services, physical therapy). Dental referrals were not captured as part of this study. The NP initiated the form only if a patient was referred. On the form she/he indicated the diagnosis, the reason for referral/consultation, the specialist to whom he/she was referring, type of x-ray, counseling or diagnostic procedure requested, and/or if a hospital referral were being made. The form was best completed by the NP rather than a central person because complete information could not always be determined from reading the chart. Forms were collected by one NP contact in each clinic and forwarded to the research statistician at the end of each month.

The development of the form created an interesting debate among the NPs at the two sites. The reason for referral is often a judgment call on the part of the NP and it also reflects the standards of care as well as the resources of each clinic. In this case, clinic resources and protocols were only somewhat different for each site. Consequently, it was felt that clinic resources and protocols did not significantly affect differences in referral patterns.

RESULTS

Over the three month period the two sites had 2,946 total encounters and a total of 190 referrals were made by the NPs in the two sites for a referral rate of 6.45%. UTHS referred 126 of the 2,039 patients (6.17% referral rate) seen in clinic for 84 different types of conditions. UHS referred 63 of 907 patients (7.05% referral rate) seen in clinic for 45 different types of conditions (Table 1). Patients for whom referrals were made were diagnosed by the NPs with a variety of conditions involving most of the major body systems. Because patient diagnoses were diverse, they were tabulated according to the International Classification of Diseases ninth revision (ICD9). This classification includes 17 major categories of diagnoses and the supplemental classification of factors influencing health status and contact with health services or V codes. Specific diagnoses codes grouped according to the 17 categories are shown in Table 1. The diagnostic category with the highest frequency at UTHS was Injury and Poisoning and included diagnoses of fractures (n=3), torn cartilage (n=1), sprains and strains (n=7), scalp laceration (n=1) and contusions (n=4). At UHS the most frequent category was Symptoms, Signs, and Ill-define Conditions. This category includes diagnoses of insomnia (n=1), rash (n=1), weight gain (n=1), headache (n=3), palpitations (n=1), cough (n=1), chest pain (n=1), abdominal pain (n=5), abnormal chest film (n=1), and abnormal pap smear (n=2).

Figure 2

Table 1: WORKING DIAGNOSIS FOR REFERRED PATIENTS (N=173)

Within the remaining 16 categories, three or more patients were referred with diagnoses of the following conditions at UTHS: nevi skin lesions (n=3), tympanic membrane rupture (n=4), knee pain (n=4), tendinitis (n=4), and PPD converters (n=5). At UHS three or more patients were referred for bronchitis (n=3), pelvic pain (n=4), and sprain (n=4). The diagnoses in the V code categories for UTHS were all in the category of Persons without reported diagnosis encountered during examination and investigation of individuals and populations and included general physical examinations (n=3), pre-employment physical examinations (n=3), gynecological examinations with pap smear (n=6), and mammogram (n=2). The diagnosis in the V code category for UTS were in the category of Persons encountering health services in circumstances related to reproduction and development with one person being seen for family planning.

When making a referral the NPs were also asked to indicate the reason for making that referral (Table 2). In some circumstances there was more than one reason for referring and they were encouraged to indicate this by marking one or more of the predetermined reasons. For the 126 referrals at UTHS, a total of 179 reasons were selected and for the 64 referrals at UHS 88 reasons were selected. The number of times each reason was selected by UTHS and UHS are shown in Table 2. The overwhelming reason selected for referral was lack of clinic resources. An example of this reason for referral would be a patient referred out for a diagnostic procedure such as x-rays, mammograms, MRIs or ultrasound examinations (Table 4).

Figure 4

Table 4: DIAGNOSTIC PROCEDURES ORDERED BY NURSE PRACTITIONERS (N=43)

Figure 5

Table 3: SPECIALISTS TO WHOM PATIENTS WERE REFERRED (N=105)

Other reasons for referral at UHS included patient insisted (n=2), no response to treatment and followed by private medical doctor (n=2), nurse practitioner out of the clinic when patient arrived, and research study. Other reasons for referral at UTHS were treatment by other PCP (n=2), patient request, and after clinic hours.

The types of specialists to whom patients were referred from the two clinics were varied (Table 3). At UTHS, 65 referrals were made to specialists while UHS made 40 referrals to specialists. Across both settings, the specialties to which NPs referred the most patients were orthopedics (10%) and dermatology (9%). Although UHS referred 5 patients to the physician at the clinic, UTHS made no such referrals. Additional specialists to which NPs referred patients at UTHS were urology (n=2), oncology (n=3), cardiology (n=4), optometry (n=1), surgery (n=5), psychology (n=2), physical therapy (n=3), chiropractor (n=1), pediatrician (n=1), rheumatology (n=1), employee assistance program (n=1), primary care physician (n=2), and immunology (n=1). NPs at UHS referred patients to the following additional specialists: cardiology (n=3), chiropractor (n=1), urology (n=2), midwife (n=1), oncologist (n=1), surgeon (n=1), and primary care physician (n=2).

{image:5}

Patients were referred for a variety of diagnostic tests (Table 4). A total of 57 patients were referred from UTHS and 33 patients from UHS. As shown in Table 4, the majority of diagnostic tests for which patients were referred at UTHS were for mammograms and at UHS, colposcopy. In addition to the predetermined categories, NPs at UHS also referred a patient for a colonoscopy while NPs at UTHS referred patients for 24 hour holter monitoring (n=2), sigmoidoscopy (n=1), EMG (n=1), arthroscopy (n=1), and arterial flow study (n=1).

No referrals were made for admission to a hospital over the 3 month period. However, 3 patients at UHS and 1 patient at UTHS were referred to the emergency department. A total of 5 patients (n=3 at UTHS and n=2 at UHS) were referred for mental health counseling; 3 patients from both UTHS and UHS were referred for fitness counseling and nutrition counseling.

A total of 35 x-rays were ordered at UTHS and 12 at UHS. At UTHS, 18 chest films were obtained and 17 extremity films. At UHS, 4 chest films, 7 extremity films, and 1 back film were ordered.

DISCUSSION

This quality assessment project shows similar trends in referral patterns of NPs in two NP managed clinics in large Southern metropolitan cities. Both clinics serve student populations and one of them also serves a large managed care population as well as several occupational health contracts throughout the city.

If NPs are to be recognized as primary care providers (PCPs) and reimbursed by managed care organizations (MCO) it is important for the MCOs to understand the strengths and limitations of NP practices in regards to referral patterns. This quality assessment project goes a long way in helping to understand some of those NP referral patterns.

The results of this quality assessment project indicate that UTHS had a referral rate for diagnostic and specialists care of 6.17% and UHS a referral rate of 7.05%. The specialists to whom the NPs referred most frequently (Table 3) were orthopedists, dermatologists, gynecologists, otolaryngologists and ophthalmologists. Assuming these referrals were problems which even a qualified internist or family physician were not able to handle in his/her office then all of the referrals would not be considered as “excess referrals” (referrals over and above the amount a physician would have referred) just because the NP provided the care. However, it should not be inferred that the fewer the referrals the better the care. This may or may not be the case. Quality of care was not the issue of this study but rather referral rates from two nurse practitioner managed clinics serving students, employees and private patients.

In this case, a managed care company or group practice would not have any added costs for any of these referrals made by the NP. On the other hand, if these referrals could have been handled by the internist or family physician then costs associated with the referral would be considered extra costs for providing that care (₁₁).

Reasons for this surprisingly low referral rate may include the relatively healthy population served by these two clinics and a natural patient self selection of critically ill patients to seek out a specialist for care rather than a NP. On the other hand, this may be an indicator of the ability of NPs to deliver comprehensive primary care to the vast majority of students and employees in a university setting. In either case, MCOs may want to take note of this when considering whether or not to place NPs on their provider panels especially for care delivery in college health settings.

Another interesting finding relates to health maintenance issues. Nurse practitioners pride themselves on providing health maintenance for patients. However, in this study over the three month period there was only one referral from the two clinics for flexible sigmoidoscopy. Both clinics provide care to a student population with UHS deriving almost all of its patients from the student body. In this case the need for such a routine screening procedure seems highly unlikely. The UTHS, however, also sees patients of all ages and would be expected to have a much higher rate of referrals over a three month period for this particular procedure as well as other more acute and chronic conditions. It is unclear as to why this particular referral rate for flexible sigmoidoscopies was so low during the study period but may indicate a need to focus on this health maintenance in the future. The referral rate for routine mammograms, however, seemed more in line with appropriate screening measures for the populations served.

Considering all referrals, the UTHS rate of gynecological referrals was 6 % while UHS was 12% (Table 3). While at first this seems disproportional between the two clinics it should be kept in mind that UTHS has a Women’s Health Care Specialist who provides colposcopy services which accounts for the difference in referral rates for this service (Table 3).

The reasons for referrals among the two sites were remarkably similar. By and large, referrals were made due to lack of clinic resources (i.e. no imaging services or specialty services in clinic). UTHS indicated this in 39% of the referrals and UHS indicated it in 46% of the referrals. Lack of skills (UTHS 22% and UHS 20%), outside the NP scope of practice (UTHS 20% and UHS 16%) and lack of knowledge (UTHS 17% and UHS 18%) all seemed less important in terms of reasons for referral rather than the lack of resources.

The literature is not helpful in terms of providing a benchmark for referral patterns of NPs. Further practice data including descriptions of referral patterns, types of health care problems seen, utilization of lab and x- ray studies and costs of care are needed. Unfortunately, most NP managed clinics are not equipped to gather data on a large scale for study purposes. The electronic medical record (EMR) is one way with which to capture these data for practice analysis, move forward a research agenda, and facilitate an overall continuous quality improvement program for the practice. The ability to document overall practice quality of care in nurse-managed clinics in the future will largely depend on the presence or absence of an EMR.

Three years ago the UTHS installed an EMR system (Practice Partners from Physician Micro Systems in Seattle, WA). This particular system provides for patient scheduling, billing and clinical records. A new Windows version of the software will allow for data analysis of this type with remarkable ease. With systems like this, provider referral patterns are easily identified and can serve as indicators for needed continuing education efforts of providers as well as staffing needs by specific specialists. For example, if dermatology referrals are high, then either the NP can obtain further continuing education in dermatology issues or the clinic might want to hire a part time dermatologist to attend clinic every week or two.

The need for more sophisticated and detailed studies related to referral patterns of NPs is apparent. These future studies must necessarily focus on more specific data related to NP referrals as they relate to physician collaboration, costs associated with those referrals, and other reasons than those described here of why NPs refer patients to other resources for care.