Research Article ABSTRACT. Amanda Myhren-Bennett College of Nursing, University of South Carolina, Columbia, SC 29208, USA

Transcription

1 Quality in Primary Care (2017) 25 (3): Insight Medical Publishing Grou Research Article Research Article Adherence to Standards of Practice Treating Diabetes between Physicians and Nurse Practitioners: The National Hosital and Ambulatory Medical Care Surveys Amanda Myhren-Bennett College of Nursing, University of South Carolina, Columbia, SC 29208, USA College of Nursing, University of South Carolina, Columbia, SC 29208, USA ABSTRACT Oen Access Background: Much of the evidence of adherence to mutually agreed-uon rules for the treatment of diabetes among hysicians and nurse ractitioners comes from single clinics or registries, which leaves oen the question as to whether these findings are nationally reresentative of current ractice. Aim: To evaluate standards of ractice for treatment of diabetes among hysicians and nurse ractitioners across the United States. Design: Observational study design using large, ublicly available datasets. Methods: We used data from the National Hosital and Ambulatory Medical Care Surveys NAMCS, NHAMCS). We assessed standards of ractice (HbA1c, foot exams, retinal exams) and delivery of atient education, using the checkbox for diabetes to identify all atients. We then examined differences in treatment using multivariate logistic regression models. Results/findings: A total samle of 10,551 ambulatory and 11,546 outatient deartment (OPD) records were analyzed (unweighted counts). Patient characteristics associated with rovider adherence in both settings were identified by airwise analysis. After adjustment and assigning survey weights, care was similar between both roviders in ambulatory settings. Odds of receiving HbA1c were 2.47 times higher among nurse ractitioners in OPD after adjustment. Across both surveys, nurse ractitioners had lower odds of roviding certain forms of atient education and counseling, including diet/nutrition, health education and other education (<0.05). Conclusion: Using nationally reresentative databases for ambulatory and OPD visits, we found that hysicians were more likely to deliver atient-based education and counseling, but were similar comared to nurse ractitioners or slightly lower in the odds of delivering mutually agreed-uon treatment of diabetes. Keywords: Diabetes; Guideline adherence; Nurse ractitioner; Health care surveys How this fits in with quality in rimary care? Few oulation-based data are available on the quality of outatient care rovided by nurse ractitioners and hysicians in the US for treatment of diabetes mellitus. What do we know? Evidence that nurse ractitioners and hysicians adhere to agree uon standards of care for treatment of diabetes mellitus is mixed and often derived from single clinical settings. Nationally reresentative datasets, such as the National Hosital and Ambulatory Medical Care Surveys, can be used to assess the quality of ractitioner involvement and outcomes of care for many illnesses and diseases. What does the aer add? Adherence to standards of care for treatment of diabetes mellitus is similar among nurse ractitioners and hysicians in ambulatory care settings. In outatient emergency deartment settings, the odds of receiving HbA1c were 2.47 times higher among nurse ractitioners, whereas receit of diet/nutrition-related counselling was 0.50 times lower among these roviders. In outatient emergency deartments, individuals with diabetes mellitus are not receiving identical treatment by nurse ractitioners and hysicians.

2 177 Background It is rojected that by 2025 demand for hysician care will increase by 17% concurrently with a hysician shortage of nearly 90,000 to 100,000 [1,2]. These shortages are exected to disroortionately affect oor, rural and minority atients, articularly those with chronic diseases [3]. Exanding the number of nurse ractitioners (NP) has been roosed as one ossible solution to meet the increased demand for care while also controlling for healthcare costs [4-7]. Such exansions could reduce the effect of hysician shortage by u to 65% [3]. One imediment to exanding the number of NPs are the limitations set by state scoe of ractice regulations [3]. Limitations in scoe of ractice for NPs have been in lace as a safety measure with such limitations being suorted by the medical community [8]. In 39 states nurse ractitioners must ractice under a hysician ractice or with a collaborative agreement with a hysician [9]. Such restrictions and limitations on ractice restrict delivery of mid-level care due to the requirement of collaboration creating the need to be in close roximity of a hysician [3]. For examle, treatment of many chronic conditions, such as diabetes, otentially would be more aroriately managed with decreased hositalizations by increasing access of care by lifting scoe of ractice restrictions on NPs [8,10]. The American Diabetes Association (ADA) has set forth guidelines for the standard of care for diabetic atients. According to the ADA every diabetic atient should receive a comrehensive medical examination. The comrehensive medical exam should include a medical history, height, weight, BMI, foot examination and HbA1c test (if not erformed in the ast three months). The ADA also recommends all atients receive education on self-management [11], which should include nutrition and exercise education [12]. These guidelines for standards of care are in lace to hel better manage glycemic control. Increased glycemic control has been found to decrease comlications from diabetes [13]. These guidelines are eriodically udated and made available to roviders to rovide acceted standards of care in management of atients with diabetes. Some studies have comared adherence to the standard of care ractices for atients with a diagnosis of diabetes between NPs and hysicians [14-18]. However, evidence thus far that NP roviders imrove uon or rovide similar care comared to hysicians has been inconsistent. For examle, Condosta [14] found that NPs erformed foot insections, odiatry referrals and ohthalmology referrals more frequently than hysicians, but hemoglobin A1C goal attainment was similar comared to hysicians. It was found that NPs more frequently documented general diabetes education, nutrition education, exercise/weight education and hemoglobin A1C s than their MD counterarts did, but not with resect to foot exams or referrals to ohthalmologists. Kuo et al. [17] found that NPs and hysicians tested for LDL at similar rates, but NPs erformed eye examinations and hemoglobin A1C testing less frequently. Conlon found that NPs lowered HbA1c and glucose levels more effectively than hysicians and also rovided education at a higher level. While many studies rovide variance among roviders adherence to standards of ractice, NPs have been found to demonstrate stricter adherence to standards of care for atients with diabetes [15,18]. Variations in findings may be due in art to small study samles or the lack of characterization of treatment atterns among different atient grous. Healthcare roviders seeking to describe and act uon these findings also require diverse and oulation-wide reresentative samles from which to evaluate care ractices. Databases such as the National Ambulatory Medical Care Survey (NAMCS) and National Hosital Ambulatory Medical Care Survey (NHAMCS), offer the oortunity to exlore whether adherence to ractice atterns is evident nationally. For examle, the NAMCS catures a reresentative samle of all atients using ambulatory services of non-federally funded hysician offices whereas the NHAMCS catures a reresentative samle of all atients in emergency and outatient deartments. Investigating ractice atterns at this scoe may rovide more definitive evidence of current similarities or differences in the delivery of care and health education to atients among NP s and hysicians. Aim The aim of this study was to reort national estimates of adherence to acceted standards of care for adult atients with a diagnosis of diabetes when treated by hysicians and NPs working in ambulatory and outatient settings. Our evaluation sought to answer three questions. First, do NPs treat the same tye of atients as their hysician counterarts? Second, do NPs rovide the same diagnostic tests and recommended screenings for atients with a diagnosis of diabetes comared to their hysician counterarts? Third, do NPs rovide the same education and counseling to atients diagnosed with diabetes as their hysician counterarts? Study Design We analyzed data from 2009 to 2011 ambulatory and outatient sections of the NAMCS and NHAMCS. Both are national surveys designed to rovide annual information about the rovision and use of medical care services in the officebased hysician ractices, with resect to the NAMCS, and about atient visits to hosital outatient (OPD) and emergency deartments, with resect to the NHAMCS. Both the NAMCS and NHAMCS are cross-sectional robability samles derived from recruiting hysicians and non-hysicians to comlete atient data and medical service forms for a reresentative samle of atient visits. Samling is conducted using a multi-stage stratified robability aroach and visit weights and clustering variables are used to derive nationally reresentative annual estimates of all ambulatory, OPD and emergency deartment visits in the United States, exclusive of federal, military, and veteran affairs facilities. Information about the samling and design of the NAMCS and NHAMCS is ublically available. This study only examines atient visits to ambulatory and OPD facilities.

3 Adherence to Standards of Practice Treating Diabetes between Physicians and Nurse Practitioners: The National Hosital and Ambulatory Medical Care Surveys 178 Inclusion Criteria Patient encounter records contained in years 2009 through 2011 for all ambulatory facilities were 32,281, 31,229 and 30,872, totaling 94,382 visits. Corresonding atient data for OPD facilities for years 2009 through 2011 were 33,551, 34,718 and 32,233, totaling 100,502 visits. We assumed that each encounter reresented a different atient, although it is ossible that multile encounters could be reresentative of the same atient. However, we were unable to account for this ossibility as there are no unique identifiers for atients in either ublicly available database. Inclusion criteria for atient encounter records included: (1) discharge alive, (2) ages 18 years and older, (3) a current diagnosis of diabetes as defined using the atient record form for the question Does the atient have diabetes? and (4) the rimary rovider defined as either a hysician (MD) or nurse ractitioner/mid-wife (NP), but not both. Our inclusion criteria catured 11.1 ercent (n=10,551) of all the samled ambulatory visits and 11.5 ercent (n=11,546) of all samled OPD visits between 2009 and Study Variables and Variable Re-classification Socio-demograhic variables included: atient age, sex, race, ethnicity and insurance tye. For atient race, the original classifications of Asian Only, Native Hawaiian/ Pacific Islander Only, American Indian/Alaska Native Only and more than one race reorted were collased into a single category,. Patient insurance tyes were divided into the following categories: uninsured (includes self-ay), commercial indemnity (including worker s comensation), Medicare and Medicaid, and. Additional atient-level variables included body mass index (BMI), number of co-morbidities, diagnosis of obesity, current smoking status, revious visits, metroolitan status (MSA), as well as geograhic region. BMI was calculated manually using the atient weight and height data columns as oosed to using the rovider entered scores, thereby increasing the number of atient weight scores by 4.9 ercent. Statistical Analysis Because the NAMCS and NHAMCS use comlex survey samling design, design effects were incororated into the statistical analyses by using SAS software [ref]. Differences between means of continuous variables were examined using Student s t test, and differences in roortions of categorical variables were examined using the Rao-Scott F adjusted chisquare statistic. The weighted samle size was used to roduce all 95 ercent confidence intervals for all comarisons. Raw numbers from the survey are rovided for clarity in reorting, articularly for instances having small counts. We erformed multile logistic regressions to analyze differences in atient visits by rovider tye. Receit of nine different care ractices were analyzed: (1) HbA1c, (2) foot exam, (3) retinal exam, (4) health education ordered, (5) diet/nutrition education, (6) exercise education, (7) weight reduction, (8) other health education, and (9) referral to other hysician. Variables identified from the airwise comarisons with <0.25 were included as otential factors that would affect the association between care rovision and rovider tye. Results In the 2009 through 2011 NHAMCS and NAMCS datasets that met our inclusion and exclusion criteria, we identified 10,551 and 11,546 ambulatory and OPD visits (unweighted counts) that indicated diabetes using the checkbox. After aroriate weighting, the estimated number of visits by atients with diabetes in the United States was 355,536,392 (standard error [S.E.]: 20,234,631) in ambulatory care and 36,649,513 (S.E.: 3,299,345) in OPD setting. Nationally, these estimates reresent 14.0% and 18.0% of all ambulatory and OPD visits in those years. Univariate analyses Socio-demograhic and clinical characteristics of atients with diabetes seen by NPs and MDs as their rimary care rovider in ambulatory and outatient care settings are shown in Table 1. In the ambulatory care setting, univariate analyses that were statistically significantly different (<0.05) showed that NP s treated a larger roortion of male atients (63.0 vs. 52.6), older atients (66.4 vs. 62.9) as well as different comosition of atients according to classifications of race. NP s also treated a different comosition of atients according to the average number of visits over the revious 12 months (5.8 vs. 4.7) as well as BMI (33.1 vs. 32.5) and current smoking status (5.7 vs. 13.7). There were no statistically significant differences in atient demograhics cared for by NPs and MDs when contrasted against insurance tye, MSA designation and geograhic region. Similarly, with the excetion of the rovision of other health education, there were no statistically significant differences in the tye or frequency of care rovided by NPs and MDs to atients having insulin deendent diabetes mellitus or noninsulin deendent diabetes mellitus. In the outatient care setting, NPs treated a larger roortion of male atients (64.8 vs. 58.4), younger atients (54.8 vs. 58.6) as well as different comosition of atients according to classification of ethnicity. NP s also treated a different comosition of atients according to BMI (35.1 vs. 33.1) and current smoking status (24.8 vs. 17.9). In contrast to ambulatory care setting, atient comosition in the outatient setting differed by insurance tye and by non-msa hosital status (31.4 vs. 14.3). With the excetion of the frequency of referrals to other hysicians (24.3 vs. 16.4), there were no statistically significant differences in the tye or frequency of care rovided by NPs and MDs with resect to diagnostic tests or atient education during care. Unadjusted regression analyses Table 2A shows unadjusted regression analysis for rocess of care for diabetes treatment among atients treated in ambulatory care. Of all care rocesses, only the odds of receit of other or unclassified atient education was statistically significantly different between NPs and MDs, with the odds 0.27 smaller that

4 179 Table 1: Characteristics of atients with Diabetes Mellitus according to whether they received ambulatory or emergency deartment care from nurse ractitioners (NP) or hysicians (MD), NAMCS and NHAMCS, years NAMCS: Ambulatory Care (weighted) NHAMCS: Outatient Care (weighted) Characteristic NP (%, SEM) MD (%, SEM) NP (%, SEM) MD (%, SEM) Predetermining factors Female 18 (27.0) 5,001 (47.4) (35.2) 4,620 (41.6) Age (SEM) 66.4 (3.0) 62.9 (0.3) (1.4) 58.6 (0.5) Race < Caucasian 36 (63.1) 6,376 (60.9) 511 (71.6) 6,501 (58.7) African American 11 (31.1) 1,252 (12.1) 131 (5.7) 2,512 (25.5) 1 (0.0) 607 (4.6) 24 (6.8) 637 (4.5) Blank 4 (5.8) 2,264 (22.4) 68 (7.7) 1,162 (11.2) Ethnicity Hisanic/Latino 12 (23.9) 1,250 (10.3) 82 (5.2) 1,729 (16.4) Non-Hisanic/Lation 37 (72.1) 7,148 (68.9) 595 (83.1) 7,660 (69.3) Blank 3 (4.0) 2,101 (20.8) 57 (7.8) 1,423 (14.3) Previous visits (SEM) 5.8 (1.0) 4.7 (0.1) < 5.4 (0.6) 5.2 (0.2) BMI (SEM) 33.1 (1.1) 32.5 (0.1) (0.7) 33.1 (0.2) < Diagnosis of obesity 12 (26.7) 2,018 (19.9) (26.6) 2,188 (19.8) Comorbidities (SEM) 3.8 (0.2) 3.2 (0.0) (0.2) 3.1 (0.1) Current smokers 4 (5.7) 1,179 (13.7) (24.8) 1,319 (17.9) Enabling factors Insurance tye Private 13 (26.6) 3,632 (39.4) 222 (27.2) 2,382 (25.5) Medicare 24 (54.5) 4,799 (47.0) 242 (32.2) 4,408 (38.1) Medicaid 9 (10.0) 966 (6.9) 163 (22.8) 2,394 (18.9) Self-Pay 2 (0.7) 490 (2.5) 75 (15.0) 935 (9.5) 4 (8.2) 612 (4.2) 32 (2.8) 693 (8.0) Geograhic region Northeast 7 (23.3) 1,932 (17.8) 249 (30.6) 3,056 (31.6) Midwest 17 (24.1) 2,506 (21.2) 162 (27.3) 2,589 (25.7) South 12 (32.2) 3,356 (39.2) 236 (35.6) 3,394 (31.6) West 16 (20.3) 2,705 (21.7) 87 (6.5) 1,773 (11.1) Non-MSA 3 (13.8) 1,059 (12.3) (31.1) 871 (14.3) Need factors HbA1C 14 (37.2) 1,548 (19.4) (33.7) 1,317 (20.1) Foot exam 10 (17.0) 750 (8.5) (15.2) 1,295 (12.0) Retinal exam 2 (5.1) 396 (4.6) (4.6) 358 (4.5) Health education ordered 24 (43.5) 4,907 (49.7) (61.0) 5,559 (51.5) Diet/Nutrition education 18 (39.1) 2,067 (22.6) (23.9) 2,051 (22.8) Exercise education 13 (20.4) 1,328 (14.4) (16.9) 1033 (13.2) Weight reduction 6 (12.5) 869 (9.9) (7.5) 571 (5.9) health education 6 (8.9) 2,819 (26.7) (40.8) 4,114 (33.8) Refer to other hysician 5 (7.7) 1,161 (10.6) (24.3) 1,507 (16.4) Source: NHAMCS and NAMCS data cycles, All standard errors of the mean (SEM) corresond to weighted mean s a atient would receive other forms of education among NPs (OR 0.27; 95% CI ; 0.012). Irresective of rovider, atient-level variations in rocess of care were observed most consistently in resect to BMI and atient race/ethnicity. Table 2B shows unadjusted regression analysis for rocess of care for diabetes treatment among atients treated in OPD. In the unadjusted regression model, the odds of atient referral to another secialist were 1.91 times larger when seen by an NP comared to an MD (1.91; ; ). No other statistically significant differences by rocess of care tye between NPs and MDs were observed. Irresective of rovider, the odds of care receit were consistently larger among atients having a diagnosis of obesity with resect to Health Education, Diet/Nutrition Education, Exercise Education and Weight Reduction. Variation in care receit also varied by atient insurance status, but not consistently by one form of indemnity ayment. Adjusted regression analyses Table 3A shows adjusted regression analysis for rocess

5 Adherence to Standards of Practice Treating Diabetes between Physicians and Nurse Practitioners: The National Hosital and Ambulatory Medical Care Surveys 180 Table 2A: Unadjusted differences in rocess of care between nurse ractitioners (NP) and hysicians (MD), NAMCS years Characteristic HbA1C Foot exam Retinal exam Health Education Diet/Nutrition Education Exercise Education Weight Reduction Education Referral Provider Phsician [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] Nurse Practitioner Female Age 2.46 ( ) 1.39 ( ) ( ) 1.29 ( ) ( ) 1.14 ( ) 1.01 ( ( ) 1.04 ( ) 0.99 ( ( ) 1.12 ( ) ( ) 1.22 ( ) ( ) 1.00 ( ) 0.98 ( ( ) 0.95 ( ( ( ) Race Caucasian [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] African American Blank Ethnicity Hisanic/ Latino Non- Hisanic/ Lation Previous visits BMI Current smokers 0.95 ( ) 1.20 ( ) 0.72 ( ) ( ) 1.67 ( ) 0.90 ( ) ( ) 0.80 ( ( ) ( ) 1.56 ( ) 0.82 ( ) ( ) 1.21 ( ) 0.76 ( ) ( ) 1.15 ( ) 0.84 ( ) ( ) 1.24 ( ) 0.90 ( ) ( ) 1.56 ( ) 1.03 ( ) ( ) 0.99 ( ( ( ) 0.94 ( ) [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] 1.18 ( ) 0.98 ( ( ) ( ) 1.01 ( ( ) ( ) 0.91 ( ) 0.99 ( ( ) ( ) 1.02 ( ( ) ( ( ( ( ) ( ) 1.04 ( ( ( ) 0.98 ( ( ) 0.96 ( ) ( ) 0.86 ( ) ( ) 1.12 ( )

6 181 Table 2B: Unadjusted differences in rocess of care between nurse ractitioners (NP) and hysicians (MD), NHAMCS outatient deartment (OPD) years Characteristic HbA1C Foot exam Retinal exam Health Education Diet/Nutrition Education Exercise Education Weight Reduction Education Referral Provider Physician [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] Nurse Practitioner Female Age 2.02 ( ) 1.27 ( ) ( ) 1.38 ( ) ( ) 1.06 ( ) ( ) 0.94 ( ) ( ) 1.06 ( ) 0.98 ( ( ) 1.00 ( ) ( ) 0.90 ( ) 0.98 ( ( ) 0.99 ( ) xx (xx - xx) Race Caucasian [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] African American Blank Ethnicity Hisanic/ Latino Non- Hisanic/Latino 0.85 ( ) 0.86 ( ) 0.61 ( ( ) 0.69 ( ) 0.55 ( ) ( ) 1.19 ( ) 0.57 ( ) ( ( ) 0.48 ( ) ( ) 0.79 ( ) 0.33 ( ) ( )* 0.50 ( )* 0.25 ( ) ( ) 0.65 ( ) 0.44 ( ) (0.53- * 1.78 ( ) 0.67 (0.32- * ( ) 1.64 ( ) 1.01 ( ( ) [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] 0.96 ( ) ( ) ( ) ( ) ( ) Insurance tye Private [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] Medicare Medicaid Self-Pay 0.94 ( ) 0.61 ( ( ) 0.77 ( ) ( ) 1.51 ( ) 1.37 ( ) 2.27 ( ) ( ) 1.54 ( ) 0.72 ( ) 2.54 ( ) ( ( ) 1.48 ( ( ) ( ( ) 1.47 ( ) 0.47 ( ) Metroolitan Status MSA [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] Non-MSA Diagnosis of obesity Previous visits BMI Current smokers 1.44 ( ) 1.26 ( ) 0.98 ( ( ) ( ) 1.79 ( ) 0.98 ( ( ) ( ) 1.73 ( ) 0.92 ( ( ( ) ( ) 1.68 ( ) 1.02 ( ( ) ( ) 2.04 ( ) 1.03 ( ( ( ) 0.64 ( ) 1.48 ( ( ) 0.20 ( ) 2.58 ( ) 2.20 ( ) 0.97 ( ( ( ) ( ) 0.89 ( ) 0.89 ( ) 1.11 ( ) 0.17 ( ) 1.03 ( ) 8.14 ( ) 0.93 ( ) 1.09 ( ) 0.80 ( ) ( ) 0.89 ( ) 1.00 ( ) 0.97 ( ) 0.53 ( ( ) 1.11 ( ) 1.02 ( ( ( ( ) 0.92 ( ) 0.93 ( ) 1.09 ( ) 1.22 ( ) 1.19 ( ) 0.97 ( ( ) 0.95 ( ) 1.01 ( ( )

7 Adherence to Standards of Practice Treating Diabetes between Physicians and Nurse Practitioners: The National Hosital and Ambulatory Medical Care Surveys 182 Table 3A: Adjusted differences in rocess of care between nurse ractitioners (NP) and hysicians (MD), NAMCS years Characteristic HbA1C Foot exam Retinal exam Health Education Diet/Nutrition Education Exercise Education Weight Reduction Education Referral Provider Physician [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] Nurse Practitioner Female Age 0.24 ( ) 1.16 ( ) ( ) 1.48 ( ) 1.0 ( ( ( ) ( ) 1.07 ( ) ( ) 1.2 ( ) ( ) 1.22 ( ) ( ) 1.23 ( ) ( ) 0.96 ( ) ( ) Race Caucasian [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] African American Blank Ethnicity Hisanic/ Latino Non- Hisanic/Latino Previous visits BMI Current smokers 0.89 ( ) 0.98 ( ) 0.83 ( ) ( ( ) 0.24 ( ) ( ) 1.4 ( ) 0.72 ( ) ( ( ) 0.56 ( ) ( ) 1.57 ( ) 0.41 ( ) ( ) 1.28 ( ) 0.35 ( ) ( ) 2.20 ( ) 0.73 ( ) ( ) 2.18 ( ) 0.74 ( ) ( ) 1.38 ( ) 0.79 ( ) 0.49 ( ) [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] 0.76 ( ) 0.97 ( ( ) ( ) 0.98 ( ( ) ( ) 0.94 ( ) 1.00 ( ( ) ( ) 0.98 ( ( ( ) ( ) 0.97 ( ( ( ) ( ) 0.99 ( ( ( ) ( ) 0.98 ( ( ) 0.96 ( ) ( ) 1.00 ( ( ) ( ) 1.02 ( ) 1.10 ( )

8 183 Table 3B: Adjusted differences in rocess of care between nurse ractitioners (NP) and hysicians (MD), NHAMCS outatient deartment (OPD) years Characteristic Provider HbA1C Foot exam Retinal exam Health Education Diet/Nutrition Education Exercise Education Weight Reduction Education Referral Physician [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] Nurse Practitioner Female Age Race 2.47 ( ) 1.76 ( ) ( ) 1.58 ( ) ( ) 0.76 ( ) ( ) 1.19 ( ) ( ) 1.45 ( ) ( ( ) ( ) 1.15 ( ) 0.98 ( ( ) 1.23 ( ) ( ) 0.74 (0.54- Caucasian [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] African American Blank Ethnicity Hisanic/ Latino Non-Hisanic/ Latino Insurance tye 1.71 ( ) 3.03 ( ) 0.14 ( ( ) 0.96 ( ) 0.49 ( ) ( ) 0.79 ( ) 1.66 ( ) ( ) 2.86 ( )* 0.55 ( ) ( ) 1.28 ( ) 0.18 ( ) ( ( ) 0.11 ( ) ( ) 1.18 ( ) 0.10 ( ) ( ) 3.97 ( ) 0.78 ( ) ( ( ) 2.55 ( ) 0.50 ( ) [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] 0.43 ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) Private [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] Medicare Medicaid Self-Pay Metroolitan Status 1.25 ( ( ) 1.77 ( ) 1.96 ( ) ( ) 1.36 ( ) 1.54 ( ) 6.09 ( ) ( ) 2.35 ( ) 0.30 ( ) 1.66 ( ) ( ) 1.27 ( ) 1.13 ( ) 0.49 ( ) ( ) 1.09 ( ) 0.92 ( ) 0.73 ( ) ( ( ) 0.83 ( ) 0.28 ( ) ( ) 0.72 ( ) 0.78 ( ) 0.09 ( ) ( ) 1.01 ( ) 0.83 ( ) 0.26 ( ) ( ) 1.26 ( )* 0.96 ( ( ) MSA [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] [reference] Non-MSA Diagnosis of obesity Previous visits BMI Current smokers 1.57 ( ) 1.77 ( ) 0.97 ( ( ( ) ( ) 2.71 ( ) 0.98 ( ( ( ( ) 3.05 ( ) 0.92 ( ( ( ) ( ) 1.94 ( ) 1.27 ( ) ( ) 1.82 ( ) 0.74 ( ) ( ) 1.56 ( ) 0.97 ( ( ) ( ) 5.81 ( ) 0.92 ( ) 1.04 ( ( ) ( ) 1.23 ( ) 0.77 ( ) ( ) 0.96 ( ) 1.02 ( ( )

9 Adherence to Standards of Practice Treating Diabetes between Physicians and Nurse Practitioners: The National Hosital and Ambulatory Medical Care Surveys 184 of care for diabetes treatment among atients treated in ambulatory care. The ambulatory model is adjusted for atient age, sex (using males as a reference), race (using whites as a reference), ethnicity (using Hisanic/Latino as a reference), number of revious visits, smoking status (using non-smokers as a reference), and BMI. After adjustment, receit of Education remained lower among NPs than MDs (0.18; ; 0.014). Health Education delivery was also 0.27 times lowers among NPs than the odds for an MD (0.27; ; 0.007). Irresective of rovider, atients with other indemnity lans exhibited consistently higher odds of receiving all classified forms of Education. No other attern of rocess of care delivery was consistently observed according to atient clinical or demograhic characteristics after adjustment. Table 3B shows adjusted regression analysis for rocess of care for diabetes treatment among atients treated in OPD. The OPD model is adjusted for atient age, sex, race, ethnicity, insurance tye (using rivate insurance as a reference), number of revious visits, non-msa (using MSA designated yes as a reference), smoking status, obesity (using no as a reference), and BMI. After adjustment, the odds of receiving HbA1c were 2.47 times larger than the odds when treated by an MD (2.47; ; 0.049). The odds of receiving Diet/Nutrition counseling were 0.50 times smaller among NPs than among MDs (0.50; ; ). Irresective of rovider, the odds of examination or atient-based education were consistently larger among atients who were diagnosed with obesity. Processes of care atters were not consistently observed across other clinical or demograhic characteristics after adjustment. Discussion Previous studies have commented on the under-adherence to standards of care among NPs with resect to agree uon treatment standards for diabetes mellitus [17], while other studies have also identified increased adherence to the same standards [15,16,19]. The majority of studies, however, suggest that NPs can safely and effectively substitute for hysicians for the treatment of diabetes [14,18,20]. However, much of the evidence for these findings comes from single clinics or hosital registries, which leaves oen the question as to whether these findings can be considered nationally reresentative of current ractice atterns. The resent study examined the function of NPs using nationally reresentative surveys from ambulatory and outatient emergency deartment visits. It also examined variations in treatment to atient-based educational and counseling to these same atients. Finally, it attemted to relate adherence to agree uon standards of care with resect to the clinical, demograhic, and geograhic rofile of the atient as well as the setting where care was delivered. Annually, there are aroximately over 118 million ambulatory and 12 million OPD visits among adults with insulin deendent diabetes mellitus and noninsulin deendent diabetes mellitus catured within the NAMCS and NHMACS registries, resectively. The NAMCS is a nationally reresentative samle covering office-based ractice of non-federally emloyed hysicians whereas the NHAMCS-OPD covers nonfederal hosital outatient deartment visits. With some imortant excetions, the odds of receiving either agreed uon standards of care or atient-based education relevant to diabetes-related care is similar between NPs and MDs when assessed nationally using both datasets. However, there are some imortant distinctions with resect to how NP ractice atterns are catured within each survey. For examle, the NAMCS samles hysicians as oosed to institutions whereas the NHAMCS is secific to institutions. In addition, the NAMCS catures NPs who work alongside hysicians, but not NPs who have their own grou ractice. As such, the NAMCS likely under-estimates NP workload [21]. With this in mind, the findings generated from the NHAMCS that odds of adherence to standard ractices of care are higher among NPs than MDs with resect to HbA1C likely rovides a more nationally reresentative samle of NP care than does the NAMCS. In this vein, our findings suort the results of revious studies also showing NPs have similar, if not better, adherence to care when comared to their hysician counterarts, and do so using evidence that more likely characterizes the conditions that are occurring across the country. With resect to atient-based education and counseling, there were some noticeable differences in ractice atterns between roviders. In the ambulatory setting, hysicians were more likely to rovide general health education and other education. Physicians were more likely to rovide diet/nutrition education in the OPD. These were unexected findings due to the fact that it is exected that atient education is emhasized throughout nursing education [16]. It is also unexected for this atient oulation because the nurse ractitioners in this study were roviding care to more comlex atients, resuming that more comlex atients would need more comrehensive education and counseling. Whether the difference may not be in the actual education or counseling rovided, but rather in the documentation of education and counseling is unknown. Nursing education has always focused on a holistic aroach [2] with an emhasis on atient education, individualized care and oen communication [22]. Our findings shows disconnect between nurse ractitioner education and training with actual ractice atterns. These differences should be investigated further with resect to other outcomes and disabilities. If these findings are robust, it would lend evidence in suort of widening rogram education and training in order to determine the ga in education to ractice. Once the ga is identified, education and training could be secially tailored so that nurse ractitioners are trained and comfortable roviding atient education. Our findings also show that atients diagnosed as obese were more likely to receive foot exams, retinal exams, general health education, diet/nutrition education, and weight reduction education. The rovision of diabetic related diagnostic tests for obese atients was not unexected. Obesity frequently leads to the develoment of other comorbid conditions such as diabetes, coronary artery disease, hyertension, stroke, cancer, etc. [23,24]. Obesity and diabetes mellitus are commonly highly correlated conditions. It could be exected to have more closely monitored diagnostic tests for the obese oulation due to the

10 185 increased likelihood of comlications from diabetes related to increased weight. This study rovides imortant and new evidence suorting the use of nurse ractitioners as rimary care roviders; further research is needed to evaluate the health outcomes of atients cared for by nurse ractitioners. Treatment of chronic conditions could be met through exanding access to mid-level care. Perhas evidence showing similar or imroved health outcomes from the ractices of nurse ractitioners would finally lead to increased scoe of ractice in the states roviding reduced and restricted ractices for nurse ractitioners. Study Limitations This study had several limitations. First, the inclusion criteria, while necessary to generate the atient oulation, greatly reduced the number of atients seen by only nurse ractitioners. Patients not included in this study may have been seen by multile roviders including a nurse ractitioner. Second, there is no way to control for the ossibility of a atient s reference or choice in care roviders. Most studies find no difference in atient reference for rovider [25,26] or increased reference for NPs [22,27,28]. Third, differences found in the education rovided to atients from nurse ractitioners comared to hysicians may be due to atients having revious visits with the same rovider. NPs treated atients with a higher number of revious visits. There is the ossibility that the atient had received education at a revious visit, exlaining why they did not receive education at the surveyed visit. Conclusion Overall, nurse ractitioners had similar ractice atterns with adherence to agree uon standards of ractice in diabetes care to their hysician counterarts. The results suort the use of nurse ractitioners as rimary care roviders for atients with diabetes. Nurse ractitioners are cometent to care for comlex atients including those with a diagnosis of diabetes. The use of nurse ractitioners will alleviate the increasing hysician shortage and holds the otential to decrease cost while imroving atient health outcomes in the rimary care setting. Further research is needed to discern differences among studies in ractice atterns between nurse ractitioners and hysicians. Further research also is needed to evaluate the ga in educating nurse ractitioners on atient education and their rovision of atient education in the clinical setting. REFERENCES 1. Commins J. Physician shortage could hit 90,000 by Health Leaders Media 2015; 34: Carlson S, Carlson W. PA: What is the difference? South Dakota Medicine 2014; 67: Xue Y, Ye A, Brewer C, Setz J. Imact of state nurse ractitioner scoe-of-ractice regulation on health care delivery: Systematic review. Nurs Outlook 2016; 64: Freeman G. Workflow changes could relieve rimary care hysician workflow. Health Leaders Media 2015: Pericak A. Increased autonomy for nurse ractitioners as a solution to the hysician shortage. Journal of New York State Nurses Association 2011; 42: Stasa H, Cashin A, Buckley T, Donaghue J. Advancing advanced ractice: Clarifying the concetual confusion. Nurse Educ Today 2014; 34: Reagan PB, Salsberry PJ. The efects of state-level scoeof-ractice regulations on the number and growth of nurse ractitioners. Nurs Outlook 2013; 61: Oliver G, Pennington L, Revelle S, Rantz M. Imact of nurse ractitioners on health outcomes of medicare and medicaid atients. Nurs Outlook 2014; 62: Hoke KaH S. Exanding access to care: Scoe of ractice laws. J Law Med Ethics 2017; 45: Stellefson M, Dinarine K, Stoka C. The chronic care model and diabetes management in the US rimary care settings: A systematic review. Prev Chronic Dis J 2013; 10: ADA. Standards of medical care in diabetes care J Clin Al Res Educ 2017; 40: Zazowsky D, Bolin J, Gaubeca VB. Handbook of Diabetes Mangement. Sringer, New York, NY Peterson KA. Diabetes management in the rimary care setting: Summary. Am J Med 2002; 113: Condosta D. Comarison between nurse ractitioner and MD roviders in diabetes care. J Nurse Pract 2012; 8: Ohman-Strickland P, Orzano A, Solberg L, DiCiccio-Bloom B, O'Malley D, et al. Quality of diabetes care in family medicine ractices: Influences of nurse-ractitioners and hysicians's assistants. Ann Fam Med 2008; 6: Lemmhslss JL. Diabetes care roceses and outcomes in atients treated by nurse ractitioners or hysicians. Diabetes Educ 2002; 28: Kuo YFG, James S, Nai-Wei C, Lwin Kyaw K, Jacques B, et al. Diabetes mellitus care rovided by nurse ractitioners vs rimary care hysicians. J Am Geriatr Soc 2015; 63: Lutifiyya MN, Tomai L, Fogner B, Cerra F, Zismer D, et al. Does rimary care diabetes management rovided to medicare atients differ between rimary care hysicians and nurse ractitioners? J Adv Nurs 2016; 73: Conlon P. Diabetes outcomes in rimary care: Evaluation of the diabetes nurse ractitioner comared to the hysician. Prim Health Care 2010; 20: Swan M, Ferguson S, Chang A, Larson E, Smaldone A. Quality of rimary care by advanced ractice nurses: A systematic review. Int J Qual Healthc 2015; 27: Morgan PA, Strand J, Ostbye T, Albanese MA. Missing in action: Care by hysician assistants and nurse ractitioners in national health surveys. Health Serv Res 2007; 42:

11 Adherence to Standards of Practice Treating Diabetes between Physicians and Nurse Practitioners: The National Hosital and Ambulatory Medical Care Surveys Budzi D, Lurie S, Singh K, Hooker R. Veterans' ercetions of care by nurse ractitioners, hysician assistants and hysicians: A comarison from satisfaction surveys. J Am Acad Nurse Pract 2010; 22: CDC. Overweight and obesity: Adult obesity causes and consequences. Centers for Disease Control and Prevention Hruby A, Manson JE, Lu Q, Malik VS, Rimm EB, et al. Determininatns and consequences of obesity. Am J Public Health 2016; 106: Cross S, Kelly P. Access to care based on state nurse ractitioner ractice regulation: Secondary data analysis results in the Medicare oulation. J Am Assoc Nurse Pract 2015; 27: Laurant M, Hermens R, Brasenning J, Akkermans R, Sibbald B, et al. An overview of atients' reference for and satisfation with, care rovided by general ractitioners and nurse ractitioners. J Clin Nurs 2008; 17: Houweling ST, Kleefstra N, Hateren KJJ, Kooy A, Goenier KH, et al. Diabetes secialist nurse as main care rovider for atients with tye 2 diabetes. Neth J Med 2009; 67: Houweling ST, Kleefstra N, Hateren A, Goenier Meyboomde Jong B, Bilo HJG. Can diabetes management be safely transferred to ractice nurses in a rimary care setting? A randomised controlled trial. J Clin Nurs 2011; 20: 9. ADDRESS FOR CORRESPONDENCE:, University of South Carolina, 1601 Greene St., SC 29208, USA, Tel: ; nathaniel@sc.edu Submitted: June 11, 2017; Acceted: June 20, 2017; Published: June 27, 2017