Evaluation and Treatment of Obesity in Primary Care

Fig. 19.1

An algorithm for identifying an appropriate weight loss option. After treating cardiovascular disease (CVD) risk factors and assessing patients’ activation for weight loss, primary care providers (PCPs) may elect to offer behavioral counseling themselves (with or without pharmacotherapy) or to provide collaborative care with other health professionals. Alternatively, PCPs may refer patients to community programs (e.g., Weight Watchers) or to obesity treatment specialists (e.g., medically supervised programs, bariatric surgery). From Tsai AG, Wadden TA. Treatment of obesity in primary care practice in the United States: a systematic review. J Gen Intern Med. 2009;24(9):1073–1079. Reprinted with permission from Springer

PCPs should consider what they and their practice are able to offer patients in terms of options for obesity treatment. As shown in Fig. 19.1, they have the option to provide behavioral treatment to patients in their offices (e.g., as outlined by the new CMS benefit). Behavioral treatment can be provided in combination with pharmacotherapy, which approximately doubles the weight loss achieved [25–27]. PCPs can also enlist other staff in the practice to provide counseling. These individuals may include nurses, medical assistants, or even well trained clerical staff [28, 29]. If treating obesity is not feasible for the providers or their practice staff, practices can contract to enlist outside clinicians to provide counseling, such as registered dietitians (RDs) or behavioral psychologists. In a 2009 review [21], Tsai and Wadden referred to this model (PCP supervising treatment, other practice staff providing counseling) as “collaborative obesity treatment within primary care.” Regardless of who does the counseling, patients are treated at their usual site of care, which has the advantage of integrating obesity treatment with other primary health care services. Due to limitations of space, costs of additional staff, or increased patient visits, provision of care in the practice itself may not be feasible, even with a RD or psychologist brought in as a consultant. PCPs who practice in larger, integrated health systems (e.g., accountable care organizations [ACOs]) are likely to have options to refer their patients for treatment outside of the practice (but within the same health system).

If treatment in the practice or in the larger health system is not an option, providers should consider referrals to obesity treatment programs in the community, or to obesity treatment specialists. Community treatment programs include nonprofit and commercial programs such as TOPS (Take Off Pounds Sensibly), Weight Watchers, Jenny Craig, and Nutrisystem. Alternatively, they may refer patients to obesity treatment specialists, such as dietitians in private practice, internists who specialize in prescribing new weight loss agents or who direct medically supervised weight loss programs (e.g., liquid meal replacement diets), or bariatric surgeons. Wherever a patient pursues treatment, the PCP’s role is to encourage continued efforts at long-term behavior change and participation in treatment, and to monitor the status of co-morbid conditions.

Supporting Evidence for Models of Intervention

Each section below examines several key studies that describe the efficacy of a specific model of treatment. The discussion is mostly limited to studies that recruited patients from primary care, using providers who were mostly naïve to treating obesity, or to studies that were modeled directly after a typical primary care environment (e.g., short duration of visits). We believe that the studies described below provide the most realistic estimate of the weight losses that can be achieved in busy primary care settings. We do not discuss trials of high-intensity treatment that were conducted in academic medical centers or community settings, as these have been reviewed previously [7, 16, 30, 31].

Studies of Weight Loss Counseling Conducted by PCPs

At least six randomized trials [32–37] have tested the effects of PCPs themselves delivering behavior weight loss counseling to their own patients. In one study, Martin and colleagues studied the effect of providing brief monthly counseling to patients in two primary care internal medicine practices [35]. Study participants were mainly African-American, low-income women, with a mean age of 41.7 years and BMI of 38.8 kg/m². Counseling visits were brief (15 min each). After 6 months, patients randomized to counseling lost more weight than those assigned to the control condition (1.4 vs. 0.3 kg; p = 0.01). However, after 18 months, the difference in weight was no longer significant (see Table 19.1).

Table 19.1

Studies of brief primary care provider (PCP) counseling and PCP counseling plus pharmacotherapy

Study	N	Interventions	Number of visits	Months	Weight change (kg)	Attrition (%)^*
Brief PCP counseling
Martin et al. [35]	144	(1) Usual care	0	18	+0.1^a	23
Martin et al. [35]	144	(2) Usual care + PCP counseling	6	18	−0.5^a	44
Christian et al. [36]	310	(1) Quarterly PCP visits	4	12	+0.6^a	15
Christian et al. [36]	310	(2) Quarterly PCP visits + PCP counseling	4	12	−0.1^a	9
Christian et al. [37]	279	(1) Usual care	1	12	+0.15^a	5
Christian et al. [37]	279	(2) Usual care + PCP counseling	2	12	−1.5^b	6.5
Ockene et al. [34]	1,162	(1) Usual care	3.4	12	0.0^a	42
		(2) PCP training	3.1	12	−1.0^a,b	42
		(3) PCP training + office support	3.6	12	−2.3^b	37
Cohen et al. [33]	30	(1) Usual care	5.2	12	+1.3^a	Not stated
Cohen et al. [33]	30	(2) Usual care + PCP counseling	9.7	12	−0.9^a
ter Bogt et al. [32]	457	(1) Usual care	2	36	−0.5	20
ter Bogt et al. [32]	457	(2) Usual care + PCP counseling	13	36	−1.1	24
Ashley et al. [39]	113	(1) RD counseling	26	12	−3.4^a	38
		(2) RD counseling + meal replacements	26	12	−7.7^b	32
		(3) PCP/RN counseling + meal replacements	26	12	−3.5^a	34
Brief PCP counseling + pharmacotherapy
Hauptman et al. [43]	635	(1) PCP guidance + placebo	10	24	−1.7^a	57
		(2) PCP guidance + orlistat, 60 mg TID	10	24	−4.5^b	44
		(3) PCP guidance + orlistat, 120 mg TID	10	24	−5.0^b	44
Poston et al. [44]	250	(1) RD/RN counseling	13	12	+1.7^a	67
		(2) Orlistat, 120 mg TID	13	12	−1.7^b	35
		(3) RD/RN counseling + orlistat, 120 mg TID	13	12	−1.7^b	34
Wadden et al. [27]^†	106	(1) Sibutramine, 10–15 mg/day	8	12	−5.0^a	18
Wadden et al. [27]^†	106	(2) Sibutramine, 10–15 mg/day + PCP counseling	8	12	−7.5^a	19

Note: For each study, under “weight change,” values labeled with different letters (a,b) are significantly different from each other at p < 0.05; RD registered dietitian, RN registered nurse, TID three times per day

^*Attrition is defined as the percentage of participants who did not contribute an in-person weight at the end of the study. An intention-to-treat analysis was used in most studies, except for three that used a completers’ analysis [33, 34, 39]

^†This study included two additional groups, both of which included intensive group lifestyle modification. The results of these groups are not displayed here

In two studies, Christian et al. tested the effectiveness of weight loss counseling for patients with type 2 diabetes or metabolic syndrome. The majority of patients in both studies were Latino and low-income. Providers in both studies received 3 h of training, and patients in both studies completed a computer-based assessment of motivation for (and barriers to) weight change. The computer program produced a personalized report with recommendations for the patient, and the provider received a copy. In the first study of patients with type 2 diabetes (mean age 53.2 years, BMI 35.1 kg/m²), patients received quarterly visits. After 12 months, the intervention group lost 0.1 kg, while the control group gained 0.6 kg (p = 0.23) [36]. In the second study of patients with metabolic syndrome (mean age 49.6 years, BMI 34.2 kg/m²), patients were seen at baseline, 6 months, and 12 months. After 12 months, weight changes in the intervention and control groups were −1.5 kg and +0.15 kg (p = 0.002 for difference) [37].

In two other studies, Ockene et al. and Cohen et al. tested the effect of brief weight loss counseling by PCPs for patients with overweight/obesity and comorbidity (hyperlipidemia and hypertension, respectively). In both studies, randomization was done at the level of the PCP, rather than at the patient level. In the study by Ockene [34], patients had a mean age of 49.3 years and BMI of 28.7 kg/m². PCPs (n = 45) were randomized to: (1) usual care; (2) brief counseling; or (3) brief counseling with in-office support. Visits were brief (8–10 min). The office support program provided prompts to the PCPs, as well as counseling algorithms and handouts. Patients had an average of 3.1–3.6 visits during the year of the study. Patients receiving the in-office support intervention lost more weight than those in the control group (2.3 vs. 0.0 kg, p < 0.001). Weight loss in the brief counseling group (1.0 kg) was not significantly different than the other two groups. In the study by Cohen [33], family medicine residents (n = 18) were randomly assigned to provide brief monthly counseling to their patients (mean age 59.5 years, BMI 34.1 kg/m²) or to provide usual care. Patients were seen an average of 9.7 and 5.2 times, respectively, over the year of the study. Residents in the intervention group were instructed on how to counsel patients on calorie restriction and healthy eating. After 1 year, patients of intervention PCPs lost 0.9 kg, while patients of usual care PCPs gained 1.3 kg (p > 0.05; exact p value not provided). Finally, ter Bogt et al. assessed the effect of counseling by nurse practitioners (NPs, mid-level primary care providers) [32]. Patients (age 56.1 years, BMI 29.6 kg/m²) received either quarterly telephone visits (plus one in-person visit), with NPs following computerized treatment guidelines, or usual care. After 1 year, weight losses were 2.0 and 0.6 kg in the intervention and control groups, respectively (p = 0.002). After 3 years, weight losses were similar in the two groups (1.1 vs. 0.5 kg, p = 0.34).

The results of these six studies indicate that low- to moderate-intensity counseling, provided by PCPs to their own patients, is not likely to produce clinically significant weight loss. While even 1 kg of weight loss may have detectable health benefit [38], the average amount of weight loss achieved in the above studies is not likely to produce substantial health benefits [5]. The low intensity of treatment and brief duration of visits are likely factors that explain the small weight losses. If patients had been seen more frequently, as recommended, by the USPSTF, weight losses might have been larger.

The possibility of greater weight loss with more frequent visits is suggested by results of a study by Ashley et al., in which both PCP counseling and “collaborative care” were tested in the same study [39]. (Note: participants in this study were volunteers from the local area, rather than patients in the practice where the study took place.) Study participants (mean age 40.4 years, BMI 30.0 kg/m²) were randomized to: (1) group behavioral counseling, 1 h visits, every 2 weeks, delivered by RDs; (2) group behavioral counseling with provision of meal replacements (Slim-Fast); or (3) individual counseling by PCPs, 10–15 min visits, every 2 weeks, with provision of meal replacements. Study participants received the LEARN Manual [40], a behavioral weight loss workbook, which was used for counseling sessions. After 1 year, weight losses in the three groups were 3.4, 7.7, and 3.5 kg, respectively (p = 0.03 for group 2, compared to groups 1 and 3). The results of this study (and of a meta-analysis of randomized trials [41] which included the study by Ashley et al.) suggest that provision of meal replacements increases weight loss, compared to a diet of self-selected food with the same calorie target. The results also suggest that RDs are at least as effective as PCPs as weight loss counselors, although this study did not do a direct comparison.

Studies of PCP Weight Loss Counseling Plus Pharmacotherapy

Randomized trials conducted in academic center or research clinics have demonstrated that adding weight loss medication to lifestyle counseling increases weight loss [26, 42]. Three randomized trials tested the effect of adding weight loss medications, simulating brief primary care office visits with PCPs providing weight loss counseling [27, 43, 44]. In the first study, Hauptman and colleagues [43] tested brief dietary counseling with placebo, orlistat 60 mg 3×/day (over-the-counter dose), or orlistat 120 mg 3×/day (prescription dose). Study participants (mean age 42.5 years, BMI 36 kg/m²) received quarterly videotapes and written materials in addition to dietary counseling. Weight losses after 2 years were 1.7, 4.5, and 5.0 kg, respectively (p = 0.001 for the orlistat groups combined, compared to placebo).

In another study of orlistat [44], Poston et al. assigned patients (mean age of 41.0 years, BMI 36.1 kg/m²) to brief counseling (15–20 min monthly visits), orlistat 120 mg 3×/day, or brief counseling plus orlistat. Counseling was provided by nurses or RDs, using the LEARN Manual. After 1 year, both orlistat groups lost 1.7 kg, while the counseling group gained 1.7 kg (p < 0.001 for orlistat groups combined, compared to counseling).

In the third study [27], Wadden et al. randomized patients (mean age 43.6 years, BMI 37.9 kg/m²) to sibutramine, 10–15 mg/day, with eight brief visits that included lifestyle counseling provided by PCPs, or sibutramine with eight brief visits that included only weigh-in and monitoring of blood pressure and pulse. Counseling visits lasted 10–15 min each and used the LEARN Manual. Weight losses after 18 weeks were 8.4 and 6.2 kg, respectively (p = 0.05), but weight losses at 1 year were not significantly different. (Note: sibutramine was removed from European and US markets in 2010 after the publication of a study indicating that it increased the risk of cardiovascular events [45].)

The results of these three studies show that medication, when added to brief counseling visits meant to simulate a primary care office environment, does increase weight loss. Sibutramine is no longer available, and orlistat, although still available both over-the-counter and as a prescription agent, is prescribed infrequently. Trials will be needed in primary care settings of two new medications approved by the FDA in 2012, phentermine–topiramate and lorcaserin [46, 47], as well of generic phentermine, which remains the most commonly prescribed weight loss agent in the United States.

Studies of the Collaborative Model of Obesity Treatment

At least four randomized trials have tested the effect of weight loss counseling conducted in the practice [28, 48–50]. Three of these studies used medical assistants or other non-PCP practice staff as counselors [28, 48, 49], and the fourth study used a RD and a fitness instructor who contracted with the practice to deliver counseling [50].

In the first study, Tsai et al. [28] trained medical assistants at two primary care practices to serve as weight loss coaches. Patients (mean age 49.5 years, BMI 36.5 kg/m²) were randomized to: (1) quarterly PCP visits and printed weight loss handouts; or (2) PCP visits plus handouts, plus eight brief counseling sessions (15–20 min each) with a weight loss coach. The written materials used for the counseling sessions were adapted from the Diabetes Prevention Program [39]. After 6 months, weight losses were 0.9 and 4.4 kg, respectively (p < 0.001), but as shown in Table 19.2, differences at 1 year were no longer significant. (Treatment was provided only during the first 6 months.)

Table 19.2

Studies of collaborative obesity care that included auxiliary professionals in the primary care practice

Study	N	Interventions	Number of visits	Months	Weight change (kg)	Attrition (%)^*
Tsai et al. [28]	50	(1) Quarterly PCP visits	4	12	−1.1^a	4
Tsai et al. [28]	50	(2) Quarterly PCP visits + MA counseling	12	12	−2.3^a	8
Wadden et al. [48]	390	(1) Usual care	4	24	−1.7^a	15
		(2) Brief lifestyle counseling (quarterly PCP visits + MA counseling)	28	24	−2.9^a,b	15
		(3) Enhanced brief lifestyle counseling (quarterly PCP visits + MA counseling + meal replacements/medication)	28	24	−4.6^b	12
Kumanyika et al. [49]	261	(1) Brief PCP counseling	4	12	−0.6^a	28
Kumanyika et al. [49]	261	(2) Brief PCP counseling + MA counseling	16	12	−1.6^a	28
Ma et al. [50]	160	(1) Usual care	3	15	−2.4^a	8.6
		(2) Adapted DPP^†	12	15	−6.3^b	8.9
		(3) Adapted DPP, self-directed	3	15	−4.5^c	7.4
Ryan et al. [51]	390	(1) Usual care	2	24	0.0^a**	55
Ryan et al. [51]	390	(2) Counseling^†† + meal replacements + medication	46	24	−8.3^b**	49

Note: For each study, under “weight change,” values labeled with different letters (a,b) are significantly different from each other at p < 0.05; PCP primary care provider, MA medical assistant, NP nurse practitioner, DPP Diabetes Prevention Program

*Attrition is defined as the percentage of participants who did not contribute an in-person weight at the end of the study. An intention-to-treat analysis was used in these studies

^†In this study, counseling was delivered by a registered dietitian who contracted with the practice, along with a fitness instructor

^††In this study, lifestyle counseling was provided by a registered dietitian, social worker, professional counselor, or marriage and family therapist

^**Weight losses represent percentage weight change, as determined by a last-observation carried forward analysis

Two larger trials have expanded the model of using medical assistants from the practice as weight loss counselors. In the first study, Wadden et al. [48] recruited 390 patients with abdominal obesity and at least 1 other component of the metabolic syndrome from six primary care practices. Study participants (mean age of 51.5 years and BMI of 38.5 kg/m²) were randomized to: (1) quarterly PCP visits and printed materials; (2) quarterly PCP visits, printed materials, plus brief monthly weight loss counseling visits, provided by a weight loss coach from the practice; or (3) all of the above interventions, plus a choice of either meal replacements or weight loss medication (orlistat or sibutramine). Weight loss coaches used written materials adapted from the Diabetes Prevention Program. After 6 months, weight losses in the three groups were 2.0, 3.5, and 6.6 kg (all significantly different from each other). After 2 years, group 3 had lost more weight than group 1 (4.6 vs. 1.7 kg, p = 0.003), but weight loss in group 2 (2.9 kg) was not significantly different from the other groups. In the second larger trial, Kumanyika et al. [49] tested the effect of using weight loss coaches from primary care practices that served primarily ethnic minority patients. Similar to the study by Wadden et al., weight loss coaches were mainly medical assistants, and they used materials adapted from the Diabetes Prevention Program. Study participants (mean age 47.2 years, BMI of 37.2 kg/m²) were randomized to PCP visits every 4 months or to PCP visits, plus brief (15–20 min) monthly visits with a weight loss coach. After 1 year, weight losses in the two groups were 0.6 and 1.6 kg, respectively (p = 0.15).

In the last of the collaborative treatment studies, Ma et al. [50] recruited patients from a single large primary care practice, all of whom had pre-diabetes by lab measurement. Study participants (mean age 59.4 years, BMI 32.0 kg/m²) were randomized to: (1) a group behavioral intervention [12 sessions], based on an adapted version of the Diabetes Prevention Program and led by a weight loss coach; (2) a self-directed intervention, using a DVD that taught patients the same curriculum in their homes; or (3) usual care. The weight loss coaches were a registered dietitian contracted to provide counseling to study participants and an exercise instructor hired to lead some groups; these two individuals were not employees of the practice. The study included a 3 month intensive intervention phase and a 12 month weight maintenance phase, during which participants in both intervention groups did not have classes but continued to receive e-mails with advice and motivational messages. After 6 months, weight losses in the three groups were 6.6, 4.3, and 0.7 kg (p < 0.001 for each comparison between groups). After 15 months, weight losses were 6.3, 4.5, and 2.4 kg, with usual care participants losing more weight and participants in the two intervention groups maintaining their weight loss (p < 0.05 for all comparisons, p < 0.001 for group behavioral intervention vs. usual care).

Together, the results of these trials suggest that obesity treatment provided by weight loss coaches from the practice (e.g., medical assistants) produces modestly greater weight loss after 1–2 years, compared to usual care. As with studies of PCP counseling, the greater weight losses in the active treatment arms of these studies are likely attributable to the greater frequency of visits (monthly with weight loss coaches, vs. approximately quarterly with PCPs). In the trial by Wadden [48], the combination of monthly counseling with a weight loss “enhancement” (meal replacements or medication) increased average weight loss to a clinically significant amount. The study by Ma et al. [50] additionally suggests the benefits of using registered dietitians, rather than medical assistants. However, controlled trials are needed to directly test this hypothesis.

The benefit of combining modalities for treatment (as was done by Wadden et al. [48]) was further highlighted by a study conducted by Ryan and colleagues [51]. They recruited patients from 7 primary care practices, where patients were covered by the same health insurance plan. Study participants (mean age 47.2 years, median BMI 46.1 kg/m²

Only gold members can continue reading. Log In or Register to continue