The Downs and Black (1998) checklist for measuring quality of both randomized and nonrandomized studies was applied to each of the included studies. This tool contains 27 items with “yes,” “no,” or “not applicable” responses to assess each study’s overall quality (10 items), external validity (3 items), study bias (7 items), confounding and selection bias (6 items), and power of the study (1 item). The creators of the tool designated a predetermined score for each item’s response that yields a potential maximum score of 30. The higher the score, the higher the quality of the study. Two investigators independently appraised each study using the checklist and then compared each item’s score in an attempt to reach a conclusive agreement of the study’s quality. If a consensus was not reached on a particular item of the checklist, a third researcher was asked to determine the final assessment. Study quality was classified as low quality/high risk of bias (score, 1–11), medium quality/unclear risk of bias (12–22), and high quality/low risk of bias (23–30).
Six studies were eligible for inclusion in the review and consisted of randomized control trials (n = 4), a cross-sectional study (n = 1), and a case study (n = 1). The most common investigated diagnoses in the included studies were Alzheimer’s dementia, diabetes, hyperlipidemia, and hypertension. All studies compared outcomes of comanagement by NP–physician teams and individual physician-led care delivery for primary care patients.
There were three outcomes that emerged from the studies: (a) PCP adherence to recommended care guidelines, (b) empirical changes in clinical patient outcomes, and (c) self-reported quality of life for the patient and their caregiver. Not every outcome was measured in every study; yet, this is the only evidence currently available surrounding NP–physician comanagement of primary care patients, and the researchers deemed it important to analyze and assess all three outcomes in order to make recommendations for practice implications and future research. The first category, adherence to recommended care, was measured as the percentage of compliance in following recommended practice guidelines that have previously been found to support optimal patient outcomes, for example, the percentage of a patient with diabetes receiving a recommended vaccination or annual ophthalmologic examination. The second category, clinical outcomes, includes changes in diagnosis-specific empirical measurements such as vital signs or laboratory values that are measured during a visit to assess how well a disease is being controlled, such as a change in blood pressure in a patient with hypertension. In patients with Alzheimer’s disease, clinical cognitive changes were assessed using the Cornell Scale for Depression in Dementia to measure changes in mood, behavior, and physical change. Behavioral change was also measured using the Neuropsychiatric Inventory (NPI), a widely used instrument in clinical trials of antidementia medications. The third category, self-reported quality of life, was measured across the studies using various instruments to scale the level of self-reported quality of life reported by patients when being treated for a particular diagnosis. The quality of life of caregivers was also evaluated. The synthesized results of each category are described below:
Adherence to recommended care guidelines
Four studies evaluated and compared how NP–physician comanagement of patient care impact compliance with completing recommended care guidelines (Ganz et al., 2010; Litaker et al., 2003; Ohman-Strickland et al., 2008; Reuben et al., 2013). In patients with hypertension, elevated cholesterol, and/or diabetes, several diagnosis-specific care guidelines are recommended to improve the quality of care delivered by the PCP in an effort to improve patient outcomes. For example, Reuben et al. (2013) found that 71% of overall recommended guidelines were completed with NP–physician comanagement compared to 35% completed by a single physician (p < .001). This study was rated by the researchers as high quality, with limited bias identified. It is important to note that there was variability in the referral of patients to NPs for comanagement of patient care, which was described as the result of patient preference or the unwillingness of physicians to make the referral.
Ganz et al. (2010) also found that a higher percentage of completed guidelines favored the NP–physician teams, specifically in patients with dementia (p < .001), falls (p = .00), incontinence (p = .01), and all diagnoses (p < .001). This study had a sample size of 200 patients and was conducted using rigorous methodology that scored as low risk of bias during quality appraisal. Litaker et al. (2003) found that recommended influenza and pneumonia vaccinations for patients with chronic disease such as diabetes was more likely to be completed by NP–physician teams (p < .001). Also, recommended smoking cessation and foot examinations were more likely to be performed (p < .001). There was no difference in the incidence of recommended completion of an annual eye exam by an ophthalmologist (p = .10). Patient education was also more likely to be completed such as dietary and activity recommendations that included sodium reduction (p < .001), moderation in alcohol consumption (p < .001), and weight control or reduction (p < .001). Although medication side effects were discussed more by NP–physician teams, there was no difference between groups in the occurrence of medication compliance conversations. This study was appraised as high quality by the researchers with a low risk of bias. They showed external validity and a clear effort to reduce bias and confounders. A small sample size (n = 156) was identified as a threat to generalizability.
Ohman-Strickland et al. (2008) examined the occurrence of recommended chronic disease assessment or monitoring specifically for diabetic patients. This was a large study across 46 practice sites and included 846 patients with diabetes. The researchers reviewed charts to determine the adherence to American Diabetes Association (2004) guidelines, such as measuring glycosylated hemoglobin (HbA1c) percentages or lipid levels that assess disease control. The PCPs in the study included NPs, physicians, and physician assistants. For the purpose of this review, we extracted only data relevant to physicians and NPs. The study found that significantly more patients were monitored for diabetic control (p < .001) and hyperlipidemia (p = .007) when comanaged by NPs and physicians compared to a single physician managing care. No difference was found for blood pressure monitoring (p = .63). This study was identified as medium quality by the researchers. There was limited information about the patients in this study, and therefore, it was hard to determine uniformity of the sample at baseline. Also, given the large amount of practices evaluated, we felt it possibly introduced bias and confounders surrounding a patient’s exposure to different team members, care processes, and facility resources. Despite the researcher’s attempt to adjust for potential confounders for organizational attributes and practices, we felt this variability could introduce influence a provider’s compliance with recommended care guidelines.
Empirical changes in clinical outcomes
Four studies examined clinical outcomes achieved by NP–physician teams compared to a single physician and were reported as empirical changes in laboratory values or vital signs. (Callahan et al., 2006; Fortinsky et al., 2014; Litaker et al., 2003; Ohman-Strickland et al., 2008). Overall, studies presented either an improvement of clinical outcomes with NP–physician comanagement or outcomes equivalent to care managed by a single physician. Two studies compared the decrease of the patient’s HbA1c percentages, used as an indicator of diabetic control, with the goal of decreasing the level to a percentage that is a recommended clinical target. In the first study, Litaker et al. (2003) reported a favorable reduction in the HbA1c level by 0.63% among patients cared for by NP–physicians teams compared to a 0.15% reduction among patients treated solely by a physician (p=.02; Litaker et al., 2003). Given the narrow window of HbA1c percentage for diabetic control, we felt this was clinically significant. This study was appraised as high quality and used rigorous methods to ensure a low risk of bias especially with attention to selection of patients by assessing their baseline medical complexity and baseline medication use.
In the second study, the attainment of HbA1c targets lacked a significant difference (p = .36; Ohman-Strickland et al., 2008). As previously mentioned, however, although this second study was much larger, the introduction of potential bias across the 46 practice sites could have potentially influenced the attainment of diabetic targets based on provider resources. In addition, there was no uniformity of patient medications prior to the start of the intervention.
These same two studies measured whether patients with hypertension and hyperlidemia achieved the intended clinical target for blood pressure (SBP < 130 mm Hg; DBP < 85 mm Hg) and lipid levels (LDL-cholesterol ≤ 100 mg/dl), and neither found a significant difference between patients comanaged and patient managed by an individual physician (blood pressure: p = .839, p = .13; lipid: p = .85, p = .78; Litaker et al., 2003; Ohman-Strickland et al., 2008). Litaker et al. (2003) did however find significantly more patients with a beneficial increase of their high-density lipoproteins (HDL) levels (p = .02).
Clinical outcomes for patients with Alzheimer’s disease were investigated in two studies using the Cornell Scale and NPI that scores cognitive and behavior impairment. The first study, Callahan et al. (2006), was rated as high quality by the researchers with low risk for bias. Their study included blinded randomization of physicians to either managing patient care single handedly or comanaging with an NP. They found no statistically significant cognitive and behavioral changes using the Cornell Scale of Depression for Dementia after 18 months (p = .94) or following a telephone interview assessment for cognition (p = .93). Using the NPI as the assessment tool, however, they found a significant increase in patient behavior (p = .01). Fortinsky et al. (2014) also used the NPI to assess patient behavior and found no statistically significant changes between the patients managed by NP–physician comanagement and physicians alone (p = .18). During quality appraisal, this study was deemed medium quality with a potential risk for bias. The sample size was extremely small (n = 31), questioning generalizability and the possibility that, because of the small sample size, there was not enough power to detect differences between groups.
Patient/caregiver quality of life
Three studies investigated self-reported patient and caregiver quality of life (Callahan et al., 2006; Fortinsky et al., 2014; Litaker et al., 2003). Caregiver quality of life was measured using two tools, NPI and Caregiver Patient Health questionnaire, and administered concurrently at either 6, 12, and/or 18 months. There was no statistical significance between groups for caregiver quality of life at 6 and 12 months, yet an improvement was noted by 18 months (p = .02; Callahan et al., 2006). Similarly, Fortinsky et al. (2014) found no statistical significance at 6 and 12 months, yet there was no 18-month follow-up assessment for comparison. Again, their small sample size and shorter follow-up compared to the other studies may have inhibited finding significance. In the same study, neither care delivery model produced a superior increase in quality of life for patient with dementia (p = .95).
A third rigorous study, appraised as high quality by the researchers, found an increase in self-reported quality of life in patients with diabetes, specifically life satisfaction, and favoring those being treated within a comanagement model (p = .04). All other quality of life measures, such as diabetes quality of life impact, social worry, or life worry, were found to be not significantly different between types of care delivery (Litaker et al., 2003).
This systematic review aimed to determine the effects of NP–physician comanagement in primary care in comparison to care delivered by an individual physician. Three outcome categories emerged from the studies: PCP adherence to recommended care guidelines, empirical changes in clinical patient outcomes, and patient or caregiver quality of life. It is evident from the literature search that, with only six studies available, the investigation of this type of care delivery is still very premature. The findings of this review, however, shed light on the promise of comanagement to help organizations meet the demand for care by adhering to recommended care guidelines and maintaining the quality of care. For example, significantly more guidelines were completed for patients with Dementia, Falls, and Incontinence. Furthermore, patients with diabetes were more likely to have their Hgba1c and lipids monitored for changes, as well as receive recommended vaccinations. The application of and compliance with these recommended care guidelines is essential for the earlier detection of disease, decreased diagnosis-specific complications, reduced hospitalizations, and reduced health care spending (Penning-van Beest et al., 2007).
Clinically, there was a variability of results for empirical patient outcomes when comparing NP–physician comanagement and a single physician managing care. For example, although some patients that were comanaged had a significantly greater reduction in target A1c levels for diabetic patients, other studies found no difference. The improvement may be attributed to the aforementioned increase in PCP adherence to recommended care guidelines given that an increased incidence of diabetic education, such as dietary and nutritional recommendations, has been found to improve patient outcomes, including HbA1c levels (Padgett, Mumford, Hynes & Carter, 1988). Yet, the lack of consistent findings across studies remains, thus suggesting that there may be differences in NP–physician comanagement interactions or processes—something worth investigating and recommended as future research. Furthermore, it is unknown whether the patients in the studies were being treated with the same medications to control total cholesterol levels and blood pressure. Only one study reported an effort to control for baseline medications but only by the number of medications and not by pharmaceutical class or mechanism of action. This lack of knowledge potentiates a variability in treatment modalities that could skew the reduction of empirical clinical changes in patients. A study that controls for baseline blood pressure, cholesterol levels, and medications is recommended.
The lack of the comanagement process or PCP interaction descriptions were evident across all studies and could have potentiated the variability of the included studies’ findings in this review. Most noteworthy, there remains a lack of instruments that assess PCP interactions with each other, and with patients or caregivers, within a comanagement model and needed to optimize the care that is delivered. For example, there was limited to no description of how NPs and physicians allocated tasks between each other or what type of ancillary support or resources were available to each PCP. As primary care continues to be delivered in team-based environments, a closer look at PCP relationships and interactions could benefit the implementation and future investigation of comanagement care delivery.
In summary, this review demonstrates promising evidence that the integration of NPs and physicians in a comanagement care model is as effective as a single physician managing a primary care patient. This finding highlights a potential solution to overcoming some of the primary care strain of managing larger patient panel sizes while maintaining the quality of care. It is important to note that individual outcome categories were not examined across all studies and were predominantly limited to two to three studies each. This further illuminates that this emerging care model requires additional research to determine its long-term effect on primary care.
The findings of this systematic review suggest significant practice, policy, and research implications. First, in terms of practice, the implementation of an NP–physician comanagement care model increases the ability of primary care organizations to adhere to recommended clinical care guidelines. It is unclear whether the increase that is shown in this review is due to an increase in the number of PCPs available to the patient or the combination of individual discipline values (medicine and advanced practice nursing) that contribute to patient care. Regardless, an increase in adherence will improve the quality of patient care delivered, as guidelines are based on the most current evidence-based and cost-effective practice. Completion of care guidelines have been found to decrease the incidence of patient disease complications, thereby alleviating unexpected exacerbations of illness that lead to increased patient visits and hospitalizations.
The process of comanagement needs to be carefully examined within individual organizational institutions for successful implementation. Characteristics of effective comanagement include effective communication, trust and respect, and a shared philosophy of care to ensure clinical alignment. These characteristics need to be supported by ensuring that both providers have access to each other’s patient care documentation, a mutually agreed up on mode of communication, and strategies to promote alignment of clinical management. Policies and an organizational culture that recognizes NPs as independent providers will be necessary to allocate equivalent support and resources that are needed to deliver optimal patient care. It has also been noted that, over time, trust and respect between providers increases and, moreover, will be strengthened when providers are given enough time and space to comanage through collaboration. Organizational policy should reflect these efforts, and the provision of such resources will promote the success of interprofessional teams when providers from various disciplines comanage the patient’s plan of care.
It is important to also recognize that NP–physician comanagement holds potential in helping organizations adhere to changes in national policy. For example, as the United States shifts from volume-based to value-based payment infrastructure, the strain to complete all recommended care guidelines is vital for an organization to be reimbursed for its patient care services. The U.S. Centers for Medicare and Medicaid Services are increasingly shifting toward reimbursements for patient care that are based on the completion of disease-targeted outcomes and guidelines that are required to maintain a high quality of care (Burwell, 2015). As previously mentioned, the ability of a single PCP to complete all recommended guidelines for each patient is often difficult and unrealistic when working as the sole PCP. This review demonstrates early evidence of the potential for NP–physician comanagement to help alleviate this organizational strain.
With regard to research implications, a substantial amount of research is warranted to understand more about the NP–physician comanagement delivery model. First, the current literature lacks a description of how NP–physician comanagement is carried out, such as the delegation of tasks, communication between the NP and physician, or how they interact. It is also unclear if each discipline is willing to work within a comanagement care delivery model. Research that is qualitative in nature is recommended to obtain descriptions of the process of existing NP–physician comanagement care delivery as well as the PCP and/or patient perspective of this type of care delivery. This should include a closer look at what attributes strengthen or impede a successful comanagement relationship between NPs and physicians so that health services researchers can continue to evaluate patient and practice outcomes for this care delivery model. Furthermore, because of some of the identified limitations of the studies included in this review, future research should include studies that control for baseline blood pressure and lab values, as well as medications, thereby eliminating bias and potential confounders when investigating comanagement effects. Finally, given that primary care is increasingly being delivered by teams with various types of PCPs, it is also recommended that future studies explore comanagement care delivery by other types of PCPs, such as physician assistants.
In summary, as the NP workforce continues to increase and policy makers encourage the expansion of NP scope of practice, such as independent NP practice, the results of this review shed light on the potential benefit of NPs and physicians comanaging primary care patients together. More research is needed to determine the best way for organizations, managers, and researchers to successfully implement such a model. The authors of this review support the movement to continually expand the NP workforce in primary care given the emerging and promising evidence of NP–physician comanagement to deliver high-quality patient care.
American College of Physicians. (2009). Nurse practitioners in primary care: A policy monograph of the American College of Physicians
. Philadelphia, PA: Author.
American Diabetes Association. (2004). Influenza and pneumonial vaccination in diabetes. Diabetes Care
, 27(1), s111–s113. doi:10.2337/diacare.27.2007.S111
Auerbach D. I. (2012). Will the NP workforce grow in the future? New forecasts and implications for healthcare delivery. Medical Care
, 50(7), 606–610. doi:10.1097/MLR.0b013e318249d6e7
Bodenheimer T., Ghorob A., Willard-Grace R., & Grumbach K. (2014). The 10 building blocks of high-performing primary care
. Annals of Family Medicine
, 12(2), 166–171.
Bodenheimer T., & Pham H. H. (2010). Primary care
: Current problems and proposed solutions. Health Affairs
, 29(5), 799–805.
Buerhaus P. I., DesRoches C. M., Dittus R., & Donelan K. (2015). Practice characteristics of primary care
nurse practitioners and physicians. Nursing Outlook
, 63(2), 144–153.
Burwell S. M. (2015). Setting value-based payment goals—HHS efforts to improve U.S. health care. New England Journal of Medicine
, 372(10), 897–899.
Callahan C. M., Boustani M. A., Unverzagt F. W., Austrom M. G., Damush T. M., Perkins A. J., … Hendrie H. C. (2006). Effectiveness of collaborative care for older adults with Alzheimer disease in primary care
: A randomized controlled trial. JAMA
, 295(18), 2148–2157.
Donelan K., DesRoches C. M., Dittus R. S., & Buerhaus P. (2013). Perspectives of physicians and nurse practitioners on primary care
practice. New England Journal of Medicine
, 368(20), 1898–1906. doi:10.1056/NEJMsa1212938
Downs S. H., & Black N. (1998). The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. Journal of Epidemiol Community Health
, 52(6), 377–384.
Fortinsky R. H., Delaney C., Harel O., Pasquale K., Schjavland E., Lynch J., … Crumb S. (2014). Results and lessons learned from a nurse practitioner
-guided dementia care intervention for primary care
patients and their family caregivers. Research in Gerontological Nursing
, 7(3), 126–137. doi:10.3928/19404921-20140113-01
Ganz D. A., Koretz B. K., Bail J. K., McCreath H. E., Wenger N. S., Roth C. P., & Reuben D. B. (2010). Nurse practitioner comanagement
for patients in an academic geriatric practice. American Journal of Managed Care
, 16(12), e343–e355.
Halcomb E. J., Davidson P. M., Daly J. P., Griffiths R., Yallop J., & Tofler G. (2005). Nursing in Australian general practice: Directions and perspectives. Australian Health Review
, 29(2), 156–166.
Lenz E. R., Mundinger M. O., Kane R. L., Hopkins S. C., & Lin S. X. (2004). Primary care
outcomes in patients treated by nurse practitioners or physicians: Two-year follow-up. Medical Care Research & Review
, 61(3), 332–351.
Litaker D., Mion L., Planavsky L., Kippes C., Mehta N., & Frolkis J. (2003). Physician–nurse practitioner
teams in chronic disease management: The impact on costs, clinical effectiveness, and patients' perception of care. Journal of Interprofessional Care
, 17(3), 223–237.
Lopez A. D., Mathers C. D., Ezzati M., Jamison D. T., & Murray C. J. (2006). Global and regional burden of disease and risk factors, 2001: Systematic analysis of population health data. The Lancet
, 367(9524), 1747–1757.
Martin-Misener R., Harbman P., Donald F., Reid K., Kilpatrick K., Carter N., … Charbonneau-Smith R. (2015). Cost-effectiveness of nurse practitioners in primary and specialised ambulatory care: Systematic review
. BMJ Open
, 5(6), e007167.
Mitka M. (2007). Looming shortage of physicians raises concerns about access to care. JAMA
, 297(10), 1045–1046.
Moher D., Liberati A., Tetzlaff J., Altman D. G. & PRISMA Group. (2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Journal of Clinical Epidemiology
, 62(10), 1006–1012. doi:10.1016/j.jclinepi.2009.06.005
Murray M., Davies M., & Boushon B. (2007). Panel size: How many patients can one doctor manage? Family Practice Management
, 14(4), 44–51.
Ohman-Strickland P. A., Orzano A. J., Hudson S. V., Solberg L. I., DiCiccio-Bloom B., O'Malley D., … Crabtree B. F. (2008). Quality of diabetes care in family medicine practices: Influence of nurse-practitioners and physician's assistants. Annals of Family Medicine
, 6(1), 14–22. doi:10.1370/afm.758
Padgett D., Mumford E., Hynes M., & Carter R. (1988). Meta-analysis of the effects of educational and psychosocial interventions on management of diabetes mellitus. Journal of Clinical Epidemiology
, 41(10), 1007–1030.
Penning-van Beest F. J., Termorshuizen F., Goettsch W. G., Klungel O. H., Kastelein J. J., & Herings R. M. (2007). Adherence to evidence-based statin guidelines reduces the risk of hospitalizations for acute myocardial infarction by 40%: A cohort study. European heart journal
, 28(2), 154–159.
Reuben D. B., Ganz D. A., Roth C. P., McCreath H. E., Ramirez K. D., & Wenger N. S. (2013). Effect of nurse practitioner comanagement
on the care of geriatric conditions. Journal of the American Geriatrics Society
, 61(6), 857–867. doi:10.1111/jgs.12268
Stanik-Hutt J., Newhouse R. P., White K. M., Johantgen M., Bass E. B., Zangaro G., … Heindel L. (2013). The quality and effectiveness of care provided by nurse practitioners. The Journal for Nurse Practitioners
, 9(8), 492–500.
Starfield B., Shi L., & Macinko J. (2005). Contribution of primary care
to health systems and health. Milbank Quarterly
, 83(3), 457–502.
Stutsky B. J., & Spence Laschinger H. (2014). Development and testing of a conceptual framework for interprofessional collaborative practice. Health and Interprofessional Practice
, 2(2), eP1066. doi:10.7710/2159-1253.1066
Wu S.-Y., & Green A. (2000). Projection of chronic illness prevalence and cost inflation
. Santa Monica, CA: RAND Health.
Yarnall K. S., Ostbye T., Krause K. M., Pollak K. I., Gradison M., & Michener J. L. (2009). Family physicians as team leaders: “Time” to share the care. Preventing Chronic Disease
, 6(2), A59.
Keywords:Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved
comanagement; nurse practitioner; primary care; systematic review; teamwork