Physician Specialty and Office Visits Made by Adults With Diagnosed Multiple Chronic Conditions: United States, 2014-2015

Abstract

Objectives

Adults with multiple chronic conditions (MCCs; ≥2 chronic conditions) account for a substantial number of visits to health care providers. The complexity of a patient’s care, including the number of chronic conditions, may differ by physician specialty. The objectives of this study were to (1) examine differences in physician office visits among adults with MCCs by physician specialty and (2) identify the types of MCC dyads (combinations of 2 chronic conditions) most common among visits to office-based physicians.

Methods

We used data from the 2014-2015 National Ambulatory Medical Care Survey (unweighted analytic sample, n = 61 682), a nationally representative survey of physician office–based ambulatory visits, to examine differences in physician office visits among adults with MCCs by physician specialty. We also identified the most commonly observed MCC dyads among these visits.

Results

During 2014-2015, 40.0% of physician office visits were made by adults with MCCs. Compared with visits for all specialties combined (40.0%), a significantly higher percentage of physician office visits among adults with MCCs were to specialists in cardiovascular disease (74.7%) and internal medicine (57.6%). For all physician specialties except psychiatry, the MCC dyads of hyperlipidemia and hypertension and diabetes and hypertension were among the most commonly observed MCC dyads among visits made by adults with MCCs.

Conclusions

Awareness of these findings may help specialists improve care for adults with MCCs. The recognition among physicians of common MCC dyads is relevant to the care management of persons with MCCs.

In 2014, 1 in 4 adults reported having received a diagnosis of multiple chronic conditions (MCCs; ≥2 chronic conditions).¹ Adults with MCCs account for a substantial number of visits to health care providers. For example, in 2009, 36.7% of office visits to US physicians were made by adults with diagnosed MCCs.² Adults with MCCs have unique health considerations that create challenges in care coordination and management.^3-10 They often receive care from multiple health care providers (a primary care provider and ≥1 specialist),¹¹ communication between patients and health care providers is often lacking,¹² health care providers do not always understand their own roles in the patient’s care,⁸ and higher morbidity can increase the number of visits to specialists.¹³

In light of these challenges, focus has been given to differences in a physician’s specialty and how physician specialty is associated with care management for adults with MCCs.^3,11,14 According to one study, clinical experience suggests that the complexity of a patient’s care and number of chronic conditions vary among medical subspecialties¹⁵; however, additional empirical evidence is needed. To our knowledge, no nationally representative studies have examined data on physician office visits among patients with MCCs by physician specialties and which dyads (combinations of 2 chronic conditions) are most common. The objectives of our study were to (1) examine differences in physician office visits among adults with MCCs by physician specialty and (2) identify the MCC dyads that are most common among adult visits to office-based physicians. This information could be used to guide future research and resources.^16,17

Methods

Data Source

We used data from the 2014-2015 National Ambulatory Medical Care Survey (NAMCS). NAMCS is an annual survey conducted by the National Center for Health Statistics of ambulatory patient visits made to nonfederal, office-based physicians in the 50 states and the District of Columbia.^18,19 Details on the NAMCS sampling design and procedures are available elsewhere.^18,19 Physicians provided verbal consent before participating. The 2014-2015 NAMCS included data on 74 042 physician office visits from 4419 physicians. In our study, we included only physician office visits made by patients aged ≥18 years, for a total of 62 196 physician office visits from 3361 physicians. We excluded 504 physician office visits made by adults to pediatricians and 10 physician office visits made by men to obstetricians/gynecologists, leaving an unweighted analytic sample of 61 682 office visits to 3224 physicians. The unweighted NAMCS physician participation rates were 43.6% for 2014 and 34.5% for 2015. We calculated these rates by dividing the number of physicians (excluding pediatricians) who provided data on at least 1 physician office visit by the total number of survey-eligible^18,19 physicians. We then multiplied this number by 100 to get the participation rate.

Measures

We identified patients as having 0, 1, or ≥2 chronic conditions. We included 16 of 20 chronic conditions identified on a list of chronic conditions developed by the US Department of Health and Human Services to promote a standardized approach to measuring MCCs²⁰: alcohol or substance abuse/dependence (includes “alcohol misuse”); Alzheimer disease/dementia; arthritis; asthma; cancer; cerebrovascular disease, history of stroke, and/or history of transient ischemic attack; chronic kidney disease or end-stage renal disease; chronic obstructive pulmonary disease (COPD); congestive heart failure; coronary artery disease, ischemic heart disease, and/or history of myocardial infarction; depression; diabetes; HIV infection/AIDS; hyperlipidemia; hypertension; and osteoporosis. As detailed elsewhere,² trained NAMCS field representatives captured data on conditions during data abstraction directly from medical records for a question asking “regardless of the diagnosis of the current visit, does the patient now have (mark all that apply). …” Four conditions (autism spectrum disorder, cardiac arrhythmias, hepatitis, and schizophrenia) included on the US Department of Health and Human Services’ standardized list of chronic conditions²⁰ were not included in our study, because the 2014-2015 NAMCS did not include these conditions in the “mark all that apply” list.

We identified physician specialty from the physician masterfiles of the American Medical Association and the American Osteopathic Association and verified specialties during the NAMCS physician interview. We grouped physicians into the following specialties^18,19: cardiovascular disease, dermatology, general family practice, general surgery, internal medicine, obstetrics/gynecology, orthopedic surgery, neurology, ophthalmology, otolaryngology, psychiatry, urology, and other (includes multiple individual physician specialties). We excluded pediatricians from the study.

Analysis

For our analyses, we used NAMCS survey weights, averaged between 2014 and 2015, to allow for national representation of office-based physician visits made by US adults. Using descriptive estimates for selected chronic conditions, number of conditions, and physician specialty, we calculated the percentage of physician office visits among adults with 0, 1, and ≥2 chronic conditions and the average number of chronic conditions. For physician office visits among adults with ≥2 chronic conditions, we compared the percentage of visits in each specialty with the percentage of visits in all specialties by using the following significance tests adjusted for the comparison of dependent samples:

$$Z=\frac{\left|X_a-X_s\right|}{\sqrt{\left(s_a^2+s_s^2\right)-2s_s^2}\left(\frac{m_s}{m_a}\right)}$$

where X_a and X_s are estimates, s_a and s_s are standard errors, and m_a and m_s are population counts of all physicians and each physician specialty, respectively. We determined significance using a 2-tailed test, with P < .05 considered significant. Finally, for each specialty, we generated estimates for physician office visits among adults with MCCs for all 120 possible MCC dyads and tabulated data on the dyads with the 5 highest percentages (similar to a previously used approach^2,21-23; again, we used significance testing adjusted for the comparison of dependent samples.

We conducted analyses by using SUDAAN version 11.0,²⁴ using survey design variables to account for covariance resulting from the NAMCS complex cluster design and subpopulation procedures (ie, SUDAAN’s nest and subpopn statements). We discussed differences in percentages for this study only if they (1) were significant at P < .05 (not adjusted for multiple comparisons) and (2) had a nondirectional Cohen’s h ≥ 0.20 (differences below this threshold are considered small).^25,26 We calculated Korn–Graubard 95% confidence intervals (CIs)²⁷ for the percentages; all percentages meet National Center for Health Statistics presentation standards.²⁸

Results

Of a total of nearly 788 million office-based physician visits, 40.0% (n = 315 million) were made by adults with diagnosed MCCs, 27.1% (n = 213 million) were made by adults with 1 chronic condition, and 32.9% (n = 259 million) were made by adults with no chronic conditions (Table 1). Hypertension (35.4%), hyperlipidemia (24.0%), and arthritis (17.8%) were the 3 most commonly observed chronic conditions among physician office visits. Most physician office visits were made to physicians specializing in general family practice (22.3%), other specialties (19.2%), and internal medicine (16.5%).

Table 1.

Percentage of diagnosed chronic conditions and physician specialties for office-based physician visits among US adults, National Ambulatory Medical Care Survey, 2014-2015^a,b

Characteristics	No. of Visits, in Thousands	%c (95% Confidence Interval)d
Total	787 676	100.0
No. of chronic conditions
0	259 051	32.9 (31.1-34.7)
1	213 208	27.1 (25.8-28.4)
≥2	315 417	40.0 (38.0-42.0)
Chronic conditionse,f
Hypertension	279 118	35.4 (33.4-37.4)
Hyperlipidemia	188 731	24.0 (22.2-25.9)
Arthritis	140 144	17.8 (15.9-19.8)
Diabetes	131 222	16.7 (15.6-17.9)
Depression	92 487	11.7 (10.8-12.7)
Cancer	60 194	7.6 (6.7-8.5)
Coronary artery disease, ischemic heart disease, and/or myocardial infarction	60 081	7.6 (6.6-8.7)
Asthma	46 729	5.9 (5.4-6.4)
Chronic kidney disease or end-stage renal disease	34 879	4.4 (3.0-6.2)
Chronic obstructive pulmonary disease	31 546	4.0 (3.6-4.5)
Alcohol or substance abuse/dependence	27 184	3.5 (3.0-4.0)
Osteoporosis	23 248	3.0 (2.7-3.4)
Cerebrovascular disease, stroke, and/or transient ischemic attack	16 298	2.1 (1.8-2.4)
Congestive heart failure	15 049	1.9 (1.5-2.3)
Alzheimer disease or dementia	7549	1.0 (0.8-1.3)
HIV/AIDS	2784	0.4 (0.3-0.6)
Physician specialtye
General family practice	175 569	22.3 (18.9-26.0)
Otherg	151 152	19.2 (15.8-22.9)
Internal medicine	129 721	16.5 (13.2-20.3)
Obstetrics/gynecologyh	62 219	7.9 (6.2-9.8)
Orthopedic surgery	49 817	6.3 (5.1-7.7)
Ophthalmology	47 706	6.1 (4.9-7.5)
Cardiovascular disease	40 054	5.1 (3.8-6.6)
Psychiatry	36 839	4.7 (3.6-6.0)
Dermatology	29 907	3.8 (3.0-4.8)
Urology	19 076	2.4 (1.8-3.1)
General surgery	17 603	2.2 (1.7-2.8)
Otolaryngology	14 894	1.9 (1.5-2.4)
Neurology	13 117	1.7 (1.2-2.3)

^aData source: National Ambulatory Medical Care Survey, 2014-2015.^18,19 Unweighted analytic sample = 61 682.

^bEstimates provided are annual averages.

^cPercentages may not total to 100.0% because of rounding.

^dConfidence intervals for certain conditions may overlap and warrant caution in determining rank order.

^eChronic conditions and physician specialties are listed in descending order of percentage of physician office visits.

^fChronic conditions are not mutually exclusive; a physician office visit may be made by an adult with >1 chronic condition. Therefore, the number of physician office visits does not equal 787 676 000.

^gDoes not include visits to pediatricians.

^hIncludes only visits by women.

We found 11 significant differences between the percentage of visits in each specialty and the percentage of visits in all specialties among adults with diagnosed MCCs. However, only 6 differences had a nondirectional Cohen h ≥ 0.20. The percentage of visits to physicians specializing in cardiovascular disease (74.7%, P < .001) and internal medicine (57.6%, P < .001) was significantly larger than the percentage of visits to physicians in all specialties combined (40.0%) (Table 2). The percentage of visits to physicians in 4 specialties was significantly lower than the percentage of visits to physicians in all specialties combined: obstetrics/gynecology (8.1%, P < .001), dermatology (21.9%, P < .001), psychiatry (23.3%, P < .001), and otolaryngology (27.3%, P < .001). The mean number of diagnosed chronic conditions among visits to physicians’ offices ranged from 2.6 to 3.2 for ≥2 chronic conditions and from 0.4 to 2.6 for any number of chronic conditions (Table 3).

Table 2.

Office-based physician visits among US adults, by physician specialty and number of diagnosed chronic conditions, National Ambulatory Medical Care Survey, 2014-2015^a,b

Physician Specialtyc	≥2 Chronic Conditionsd		1 Chronic Conditiond		0 Chronic Conditionsd
	Estimated No. of Visits, in Thousandsd	%e (95% CI)	Estimated No. of Visits, in Thousandsd	%e (95% CI)	Estimated No. of Visits, in Thousandsd	%e (95% CI)
Cardiovascular disease	29 906	74.7 f (69.4-79.5)	6606	16.5 (13.6-19.7)	3542	8.8 (6.5-11.6)
Internal medicine	74 655	57.6 f (52.3-62.8)	32 026	24.7 (21.9-27.7)	23 040	17.8 (14.6-21.3)
General family practice	81 618	46.5 (42.7-50.3)	42 417	24.2 (21.1-27.6)	51 534	29.4 (26.7-32.2)
Overall (all specialties)g	315 417	40.0 (38.0-42.0)	213 208	27.1 (25.8-28.4)	259 051	32.9 (31.1-34.7)
Otherg	58 250	38.5 (34.4-42.8)	46 559	30.8 (27.4-34.4)	46 344	30.7 (26.8-34.8)
Orthopedic surgery	16 275	32.7 (28.6-37.0)	17 563	35.3 (32.9-37.8)	15 979	32.1 (28.6-35.8)
Urology	6164	32.3 (28.3-36.5)	5589	29.3 (27.0-31.7)	7323	38.4 (33.9-43.0)
Ophthalmology	14 827	31.1 (26.2-36.3)	13 496	28.3 (23.1-34.0)	19 384	40.6 (35.8-45.5)
General surgery	5430	30.8 (25.7-36.2)	4206	23.9 (20.7-27.3)	7967	45.3 (39.0-51.7)
Neurology	4042	30.8 (23.8-38.5)	4085	31.1 (26.3-36.3)	4991	38.0 (29.7-46.9)
Otolaryngology	4073	27.3f (22.7-32.3)	3205	21.5 (18.8-24.4)	7616	51.1 (45.5-56.7)
Psychiatry	8582	23.3f (19.3-27.7)	19 764	53.6 (49.3-57.9)	8493	23.1 (19.1-27.5)
Dermatology	6538	21.9f (18.4-25.7)	7910	26.4 (22.8-30.3)	15 459	51.7 (46.4-57.0)
Obstetrics/gynecologyh	5059	8.1f (6.0-10.6)	9781	15.7 (14.1-17.4)	47 379	76.1 (72.7-79.3)

^aData source: National Ambulatory Medical Care Survey, 2014-2015.^18,19 Unweighted analytic sample = 61 682.

^bEstimates are annual averages.

^cPhysician specialties listed in descending order of percentage of visits made by adults with ≥2 chronic conditions.

^dThe following chronic conditions were included: alcohol or substance abuse/dependence; Alzheimer disease or dementia; arthritis; asthma; cancer; cerebrovascular disease, stroke, and/or transient ischemic attack; chronic kidney disease or end-stage renal disease; chronic obstructive pulmonary disease; congestive heart failure; coronary artery disease, ischemic heart disease, or myocardial infarction; depression; diabetes; HIV/AIDS; hyperlipidemia; hypertension; and osteoporosis.

^ePercentages may not total to 100.0% because of rounding, and number of visits may not add up to the total. The mean number of chronic conditions is presented in Table 3.

^fSignificant at P < .05 (2-tailed; adjusted for comparison of dependent samples) and Cohen h ≥ 0.20 compared with “overall (all specialties)” for physician office visits made by adults with ≥2 chronic conditions.

^gDoes not include visits to pediatricians.

^hOnly includes visits made by women.

Table 3.

Mean number of diagnosed chronic conditions at office-based physician visits among US adults, by physician specialty, National Ambulatory Medical Care Survey, 2014-2015^a

Physician Specialtyb	Mean No. (95% CI) of Visitsc
	≥2 Chronic Conditions (Range, 2-16)	Any No. of Conditions (Range, 0-16)
Cardiovascular disease	3.2d (3.2-3.4)	2.6d (2.4-2.8)
Internal medicine	3.2 (2.9-3.4)	2.1d (1.8-2.4)
General family practice	3.0 (2.9-3.1)	1.6 (1.5-1.7)
Overall (all specialties)e	3.0 (2.9-3.1)	1.5 (1.4-1.6)
Othere	3.0 (2.9-3.2)	1.5 (1.3-1.6)
Orthopedic surgery	2.8d (2.7-2.8)	1.3d (1.1-1.4)
Urology	2.8d (2.8-2.9)	1.2d (1.1-1.3)
Ophthalmology	2.7d (2.5-2.9)	1.1d (1.0-1.3)
General surgery	2.8d (2.7-2.9)	1.1d (0.9-1.3)
Neurology	3.0 (2.8-3.3)	1.2d (1.0-1.5)
Otolaryngology	2.8d (2.7-2.9)	1.0d (0.8-1.1)
Psychiatry	2.5d (2.4-2.6)	1.1d (1.0-1.2)
Dermatology	2.7d (2.5-2.9)	0.8d (0.7-1.0)
Obstetrics/gynecologyf	2.6d (2.5-2.8)	0.4d (0.3-0.4)

^aData source: National Ambulatory Medical Care Survey, 2014-2015.^18,19 Unweighted analytic sample = 61 682.

^bPhysician specialties are listed in descending order of percentage of visits made by adults with ≥2 chronic conditions (see Table 2). All means have a relative standard error <30.0%. SUDAAN version 11.0’s default 95% CIs are presented.²⁴

^cEstimates provided are annual means. The following chronic conditions were included: alcohol or substance abuse/dependence; Alzheimer disease or dementia; arthritis; asthma; cancer; cerebrovascular disease, stroke, and/or transient ischemic attack; chronic kidney disease or end-stage renal disease; chronic obstructive pulmonary disease; congestive heart failure; coronary artery disease, ischemic heart disease, or myocardial infarction; depression; diabetes; HIV/AIDS; hyperlipidemia; hypertension; and osteoporosis.

^dSignificant at P < .05 (2-tailed; not adjusted for comparison of dependent samples), compared with overall (all specialties).

^eDoes not include visits to pediatricians.

^fOnly includes visits made by women.

Several general patterns emerged in the percentage of physician office visits made by adults with MCC dyads (Table 4). The dyad of hyperlipidemia and hypertension had the highest observed percentage (41.1%) among physician office visits made by adults with MCCs. In addition, for 12 of 13 physician specialty groups, among physician office visits made by adults with MCCs, the dyad of hyperlipidemia and hypertension was among the top-5 most commonly observed MCC dyads. Compared with the percentage of office visits made by adults with MCCs to physicians overall (41.1%), the percentage of physician office visits made by adults with the MCC dyad of hyperlipidemia and hypertension was significantly higher among specialists in cardiovascular disease (62.2%, P < .001) and significantly lower among specialists in orthopedic surgery (18.4%), dermatology (20.7%), obstetrics/gynecology (22.3%), neurology (25.6%), urology (26.8%), general surgery (28.4%), otolaryngology (29.1%), and ophthalmology (29.7%) (all P < .001).

Table 4.

Five most commonly observed diagnosed chronic condition dyads among visits to office-based physicians among adults with ≥2 chronic conditions, by physician specialty, National Ambulatory Medical Care Survey, 2014-2015^a,b

Physician Specialty and MCC Dyadc	Estimated No. of Visits, in Thousands	% (95% Confidence Interval)
Overall (all specialties)	315 417	—
Hyperlipidemia and hypertension	129 718	41.1 (38.8-43.5)
Diabetes and hypertension	87 399	27.7 (26.1-29.4)
Arthritis and hypertension	67 803	21.5 (19.1-24.1)
Diabetes and hyperlipidemia	61 192	19.4 (18.0-20.9)
Coronary artery disease, ischemic heart disease, and/or myocardial infarction and hypertension	45 452	14.4 (12.8-16.1)
General family practice	81 618	—
Hyperlipidemia and hypertension	39 790	48.8 (44.7-52.8)
Diabetes and hypertension	24 704	30.3 (27.8-32.9)
Diabetes and hyperlipidemia	18 249	22.4 (20.1-24.8)
Arthritis and hypertension	16 052	19.7 (16.9-22.8)
Arthritis and hyperlipidemia	11 456	14.0 (11.8-16.6)
Internal medicine	74 655	—
Hyperlipidemia and hypertension	36 689	49.1 (44.0-54.3)
Diabetes and hypertension	21 711	29.1 (25.2-33.3)
Diabetes and hyperlipidemia	18 971	25.4 (22.5-28.5)
Arthritis and hypertension	17 733	23.8 (16.3-33.3)
Arthritis and hyperlipidemia	14 713	19.7 (15.9-24.1)
Other	58 250	—
Hyperlipidemia and hypertension	18 672	32.1 (27.4-37.1)
Diabetes and hypertension	16 103	27.6 (23.1-32.7)
Arthritis and hypertension	12 377	21.2 (18.2-24.7)
Diabetes and hyperlipidemia	10 761	18.5 (14.6-23.2)
Cancer and hypertension	7640	13.1 (9.0-18.8)
Cardiovascular disease	29 906	—
Hyperlipidemia and hypertension	18 589	62.2d (55.3-68.6)
Coronary artery disease, ischemic heart disease, and/or myocardial infarction and hypertension	14 458	48.3 (40.9-55.8)
Coronary artery disease, ischemic heart disease, and/or myocardial infarction and hyperlipidemia	12 492	41.8d (35.2-48.7)
Diabetes and hypertension	7856	26.3 (22.3-30.7)
Diabetes and hyperlipidemia	6077	20.3 (16.6-24.7)
Orthopedic surgery	16 275	—
Arthritis and hypertension	8165	50.2d (46.3-54.1)
Arthritis and hyperlipidemia	3253	20.0 (15.7-25.1)
Hyperlipidemia and hypertension	2992	18.4d (14.8-22.6)
Arthritis and diabetes	2961	18.2d (15.4-21.4)
Diabetes and hypertension	2639	16.2d (14.1-18.6)
Ophthalmology	14 827
Diabetes and hypertension	6111	41.2d (36.5-46.1)
Hyperlipidemia and hypertension	4399	29.7d (21.5-39.4)
Arthritis and hypertension	4341	29.3 (20.0-40.6)
Diabetes and hyperlipidemia	2375	16.0 (12.1-21.0)
Arthritis and diabetes	1938	13.1 (9.2-18.3)
Psychiatry	8582
Depression and hypertension	3027	35.3d (27.8-43.5)
Alcohol or substance abuse/dependence and depression	2820	32.9d (26.7-39.7)
Arthritis and depression	1179	13.7d (9.9-18.8)
Depression and diabetes	1025	11.9d (9.2-15.3)
Depression and hyperlipidemia	928	10.8 (7.7-14.9)
Dermatology	6538	—
Cancer and hypertension	1600	24.5d (18.2-32.0)
Diabetes and hypertension	1493	22.8 (17.1-29.8)
Arthritis and hypertension	1396	21.3 (16.8-26.7)
Hyperlipidemia and hypertension	1354	20.7d (13.7-30.1)
Arthritis and cancer	716	11.0d (7.4-15.9)
Urology	6164	—
Cancer and hypertension	1779	28.9d (24.3-34.0)
Hyperlipidemia and hypertension	1650	26.8d (22.0-32.2)
Diabetes and hypertension	1553	25.2 (20.3-30.9)
Arthritis and hypertension	952	15.4 (11.5-20.5)
Coronary artery disease, ischemic heart disease, and/or myocardial infarction and hypertension	862	14.0 (10.8-18.0)
General surgery	5430	—
Hyperlipidemia and hypertension	1542	28.4d (22.9-34.6)
Diabetes and hypertension	1267	23.3 (18.6-28.8)
Arthritis and hypertension	975	18.0 (12.4-25.3)
Cancer and hypertension	869	16.0d (12.2-20.8)
Depression and hypertension	672	12.4 (7.7-19.4)
Obstetrics/gynecology	5059	—
Diabetes and hypertension	1211	23.9d (18.6-30.2)
Depression and hypertension	1158	22.9d (18.2-28.4)
Hyperlipidemia and hypertension	1129	22.3d (15.4-31.3)
Asthma and hypertension	586	11.6 (7.4-17.6)
Arthritis and hypertension	554	10.9d (6.9-16.9)
Otolaryngology	4073	—
Hyperlipidemia and hypertension	1186	29.1d (23.1-36.0)
Diabetes and hypertension	1076	26.4 (22.0-31.4)
Cancer and hypertension	694	17.0d (13.2-21.7)
Arthritis and hypertension	666	16.4 (12.4-21.3)
Diabetes and hyperlipidemia	500	12.3 (8.3-17.9)
Neurology	4042	—
Hyperlipidemia and hypertension	1035	25.6d (18.1-34.9)
Arthritis and hypertension	972	24.0 (15.4-35.5)
Diabetes and hypertension	908	22.5 (18.5-27.0)
Depression and hypertension	719	17.8d (13.9-22.4)
Cerebrovascular disease, stroke, and/or transient ischemic attack and hypertension	452	17.4d (5.6-21.1)

Abbreviation: MCC, multiple chronic conditions.

^aData source: National Ambulatory Medical Care Survey, 2014-2015.^18,19 Unweighted analytic sample = 61 682.

^bEstimates provided are annual averages.

^cThe following chronic conditions were included: alcohol or substance abuse/dependence; Alzheimer disease or dementia; arthritis; asthma; cancer; cerebrovascular disease, stroke, and/or transient ischemic attack; chronic kidney disease or end-stage renal failure; chronic obstructive pulmonary disease; congestive heart failure; coronary artery disease, ischemic heart disease, and/or myocardial infarction; depression; diabetes; HIV/AIDS; hyperlipidemia; hypertension; and osteoporosis. Specialties are listed by descending order of total number of visits. Within dyads, chronic conditions are listed in alphabetical order. Obstetrics/gynecology specialty only includes visits made by women. Visits to pediatricians were not included in the “other” specialty or “overall (all specialties)” estimates. A single physician office visit could be made by an adult with ≥2 MCCs; therefore, dyad combinations are not mutually exclusive; therefore, we did not conduct statistical testing among dyads within a single specialty.

^dSignificant at P < .05 (2-tailed; adjusted for comparison of dependent samples) and Cohen h ≥ 0.20 compared with same MCC dyad for overall (all specialties). Note that for comparisons of an MCC dyad that was 1 of the 5 most common for a specialty but not for overall (all specialties), the percentage for overall (all specialties) is not presented.

Among office-based physician visits made by adults with MCCs, the second most commonly observed MCC dyad among all physician specialties was diabetes and hypertension (27.7%). It was also 1 of the 5 most commonly observed MCC dyads among 12 of 13 physician specialties (Table 4). Compared with the percentage of office-based physician visits among adults with MCCs to physicians overall, the percentage of physician office visits among adults with the MCC dyad of diabetes and hypertension was significantly lower among orthopedic surgeons (16.2%, P < .001) and significantly higher among ophthalmologists (41.2%, P < .001).

The MCC dyad of arthritis and hypertension had the third-highest observed percentage among physician office visits made by adults with MCCs (21.5%), and it was 1 of the 5 most commonly observed MCC dyads among physician office visits for 11 of 13 specialties (Table 4). Compared with the percentage of office-based physician visits made by adults with MCCs to physicians overall, the percentage of physician office visits among adults with the MCC dyad of arthritis and hypertension was significantly higher among specialists in orthopedic surgery (50.2%, P < .001) and significantly lower among specialists in obstetrics/gynecology (10.9%, P < .001).

We also found several MCC dyads in the top 5 percentages for physician specialty groups but not for physicians overall (Table 4). Several comparisons were both significant and had a nondirectional Cohen h ≥ 0.20. For example, the percentage of physician office visits made by adults with the MCC dyad of cancer and hypertension was significantly higher among specialists in urology (28.9%, P < .001), dermatology (24.5%, P < .001), otolaryngology (17.0%, P < .001), and general surgery (16.0%, P < .001) than among physicians overall (8.2%). The percentage of physician office visits made by adults with the MCC dyad of depression and hypertension was significantly higher among psychiatrists (35.3%, P < .001), obstetricians/gynecologists (22.9%, P < .001), and neurologists (17.8%, P < .001) than among all physicians (10.7%). The percentage of physician office visits made by adults with the MCC dyad of arthritis and diabetes was significantly higher among orthopedic surgeons (18.2%, P < .001) than among all physicians (9.4%).

The percentage of physician office visits made by adults with the MCC dyad of coronary artery disease, ischemic heart disease, and/or myocardial infarction and hyperlipidemia was 3 times higher among cardiologists (41.8%, P < .001) than among physicians overall (11.1%) (Table 4). Among neurologists, 17.4% (P < .001) of physician office visits were made by adults with the MCC dyad of cerebrovascular disease, stroke, and/or transient ischemic attack and hypertension, which was almost 14 percentage points higher than among office visits made by adults with MCCs to physicians overall (3.7%). The percentage of physician office visits made by adults with the MCC dyad of arthritis and cancer was significantly higher among dermatologists (11.0%, P < .001) than among all specialty groups combined (3.8%). We found several differences between psychiatrists and all physician specialties for physician office visits made by adults with MCCs. The percentage of physician office visits to psychiatrists was higher than the percentage of physician office visits to all specialty groups combined for the dyads of alcohol and/or substance abuse/dependence and depression (32.9% vs 2.3%; P < .001), arthritis and depression (13.7% vs 5.8%; P < .001), and depression and diabetes (11.9% vs 4.8%; P < .001).

Discussion

Our analysis found that a higher percentage of physician office–based visits were made by adults with MCCs to physicians specializing in cardiovascular disease and internal medicine than to physicians overall. In contrast, a lower percentage of physician office–based visits were made by adults with MCCs to physicians specializing in obstetrics/gynecology, dermatology, psychiatry, and otolaryngology than to physicians overall. These estimates show that about 3 of 4 visits to cardiologists and more than half of visits to internists were made by adults with MCCs, which supports the finding of other studies that multimorbidity is a common chronic condition managed in practice.^29,30

Our analysis also found similarities and differences by specialties. For all specialties examined except psychiatry, the MCC dyad of hyperlipidemia and hypertension was among the 5 most commonly observed MCC dyads for physician office visits made by adults with MCCs (not surprising based on previous findings).^2,21,31 Furthermore, among all specialties examined except psychiatry, another 1 of the 5 most commonly observed MCC dyads among physician office visits made by adults with MCCs was diabetes and hypertension (found as a common dyad in some research^2,21,23 but not others).³¹ Furthermore, the MCC dyad of arthritis and hypertension was 1 of the 5 most commonly observed MCC dyads among physician office visits made by adults with MCCs for all specialties except psychiatry and cardiovascular disease, which is consistent with studies showing that this dyad is common among adults²³ and visits made by adults aged ≥45 with MCCs² but inconsistent with estimates for Medicare beneficiaries.^21,31

Some commonly observed dyads may consist of chronic conditions that are relevant to a physician’s specialty (eg, for cardiologists, hyperlipidemia and hypertension); however, other dyads may consist of 1 condition that is directly related to a physician’s specialty and another condition that may simply be due to a high prevalence of that condition among the US adult population (eg, for orthopedic surgeons, arthritis and hypertension). Regardless, recognizing these MCC dyads as some of the most commonly observed among physician office visits made by adults with MCCs (with the exception of psychiatrists) is relevant to the general care of persons with MCCs. This recognition may be useful in professional training (eg, in such publications as the HHS Strategic Framework on MCC ^32,33) and could serve as a reference or resource for efforts focused on improving the health of persons with MCCs,³⁴ health care providers involved with patient-centered medical homes and primary care transformation where comprehensive or coordinated care are critical functions,^35,36 and the Million Hearts initiative, which focuses on blood pressure control and cholesterol management.³⁷

We also found differences in MCC dyads among the 5 most commonly observed MCC dyads among visits made by adults with MCCs for a particular specialty but not for physicians overall. For example, the cancer and hypertension dyad was 1 of the 5 most commonly observed dyads only among visits to urologists, dermatologists, otolaryngologists, and general surgeons but not among visits to physicians overall (which might not be surprising, but these estimates provide empirical evidence). Although such dyads may not generally be common among visits to all physicians, physicians in certain specialties should be aware of an increased opportunity to provide care to a patient with a combination of chronic conditions.

The finding that 40.0% of visits to office-based physicians were made by adults with MCCs updates the broader literature, which found a slightly lower percentage of visits to office-based physicians made by adults with MCCs in 2009.² This higher percentage of office-based physician visits made by adults with MCCs in 2014-2015 may be attributed to the addition of more NAMCS checkboxes in recent years that capture data on conditions that were not previously included (ie, alcohol or substance abuse/dependence, Alzheimer disease or dementia, and HIV/AIDS). This higher percentage in our study may be expected because more chronic conditions were accounted for in our study than in the previous study,² and the percentage of adults who had seen or talked to a health care professional increased with the number of diagnosed chronic conditions.³⁸ Although these findings do not directly inform care-coordination issues of providing care to adults with MCCs,⁸ they indicate that 2 specialties—cardiovascular disease and internal medicine—have a higher percentage of physician office visits made by adults with MCCs than the national percentage. For cardiologists (a medical specialty), this finding contradicts previous estimates by broader physician specialty groupings and found that primary care providers (vs surgeons and medical specialists) have the highest percentage of physician office visits made by adults with MCCs.³⁹ However, because the MCC dyads of hyperlipidemia and hypertension and diabetes and hypertension both include chronic conditions associated with cardiovascular disease,⁴⁰ the finding that cardiologists have a higher percentage of physician office visits made by adults with MCCs than the national percentage may not be a surprise.

Limitations

This study had several limitations. First, the unit of analysis for NAMCS is a patient ambulatory care visit to an office-based physician^18,19; therefore, the same adult could have had >1 visit included in this study. However, visit data are collected from a physician for only a single week, which suggests a low chance that a patient would have >1 visit selected. Second, NAMCS does not capture data on nonambulatory visits or visits made to hospitals or other non–office-based settings (although the 2014-2015 NAMCS did include data on visits to physician offices owned by hospitals). Data on such visits would not be included in the analysis. Third, we examined only 16 chronic conditions. Although these conditions were outlined in the definition of MCCs by Goodman et al²⁰ and represent more conditions than previous years of the NAMCS,² these conditions are a fraction of the range of chronic conditions among US adults. Broader measurement of certain conditions using NAMCS (eg, cancer, regardless of type or stage) could also have influenced our findings. Fourth, we examined only diagnosed conditions. As such, unless a physician recorded in a medical record that a visit was made by a patient with MCCs (which could be less likely among physicians without an integrated/interoperable health information system than among physicians with an integrated/interoperable system), this visit would not be identified as such in NAMCS. In addition, it may not be necessary for some specialists to document or have awareness of all of a patient’s chronic conditions to treat the patient. The fact that it may not be necessary for some specialists to document or have awareness of all of a patient’s chronic conditions to treat the patient may have been the case in our research, in which we observed a lower percentage of visits made by adults diagnosed with MCCs among certain specialties (eg, obstetrics/gynecology, dermatology) than among all physicians.

Fifth, our research was descriptive. Although such estimates have not previously been examined and they help fill knowledge gaps,^32,33 they do not account for additional variables such as characteristics of the patient (eg, sex, age), physician (eg, average distribution of his/her patients, subspecialty focus on geriatric care), or physician’s practice (eg, accepts Medicare, uses an electronic health record system). Further research could expand these findings by considering additional factors in multivariate analyses to explore whether they explain some variation in whether visits made by adults with MCCs differ by physician specialty. Sixth, we made multiple comparisons without adjustment in this study, which increased the possibility of type I error; however, presenting only comparisons that were both significant and had a nondirectional Cohen h ≥ 0.20 may have helped limit this possibility. Finally, although NAMCS weights were adjusted for nonresponse, participation rates may have been a limitation.

Conclusion

Our study findings complement previous research³ and are potentially the first nationally representative estimates of differences by physician specialty among visits to office-based physicians made by adults with MCCs. Our findings show that physicians with certain specialties may have a higher (or lower) percentage of their visits made by adults with MCCs. In addition, although physician office visits made by adults with certain MCC dyads (eg, hyperlipidemia and hypertension, diabetes and hypertension, and arthritis and hypertension) may be common across specialties, several specialties (eg, urology, dermatology, psychiatry) have a higher percentage of visits made by adults with MCC dyads not frequently observed in most specialties or overall. Further investigation of additional patient, physician, and practice characteristics may yield a better understanding of differences in physician office visits made by adults with MCCs and, ultimately, further inform care management.

ORCID iD

Brian W. Ward https://orcid.org/0000-0001-8257-1592