Improving Residents' Clinical Approach to Obesity
Improving Residents' Clinical Approach to Obesity
This study ran from July 2010 to June 2011 and was approved by the University of Florida (UF) Institutional Review Board (IRB).
The UF IM Residency programme is a large university-based postgraduate programme with 10 medicine subspecialty divisions, training 75 IM residents over a 3-year period (also known as categorical residents). Residents receive formal didactic education on important research and clinical topics in IM through once weekly mortality and morbidity conferences, once weekly medicine grand rounds, daily 1 h morning reports, daily 1 h resident noon conferences, and monthly 1 h conferences before continuity clinic. For experiential training, residents rotate through two large hospitals—Malcolm Randall Veterans Affairs Medical Center (VAMC), the second largest US veterans' hospital system with 239 beds; and UF Health Shands Hospital, an 850-bed, non-profit tertiary care referral centre for the south-eastern USA. Categorical residents are also assigned to a 3-year continuity clinic at one of three sites (UF Medical Plaza, UF Tower Hill, or VAMC) where they provide ambulatory care to a longitudinal panel of primary care patients one half-day per week.
Physician-participants included all postgraduate year 1 (PGY-1), PGY-2, and PGY-3 categorical IM residents in the 2010–2011 academic year who had managed a panel of their own continuity clinic patients for at least 4 months before and 6 months after the multidisciplinary obesity-specific didactic sessions (MODS) curriculum. Being in a 3-year postgraduate training programme, these residents were present for our entire educational intervention.
Medical records of patients with body mass index (BMI) >25 kg/m in the three IM residents' clinic sites were reviewed retrospectively for clinical outcomes and residents' clinical practice behaviours related to obesity management. The patient population in our large urbanised academic clinic consisted of a 1.5 :1 female : male ratio with average age of 52 years old, a variety of chronic medical conditions (ie, coronary artery disease, chronic obstructive pulmonary disease, diabetes), and a sizeable population of vulnerable, indigent patients (35% non-Caucasian, 13% Medicaid, 10% uninsured). Patient eligibility included age >18 years, BMI >25 kg/m, and established care in the IM residents' clinics at least 4 months before and 6 months after the MODS intervention. The only exclusion criterion was pregnancy.
The educational intervention consisted of four 1 h obesity-specific didactic sessions scheduled during the IM resident noon conference every 2 weeks over an 8-week time period from 5 November 2010 to 17 December 2010 (figure 1). These lectures were delivered by a multidisciplinary team of academic expert lecturers with extensive research and clinical training in obesity medicine. The team included an endocrinologist, a psychiatrist, a public health physician, a nutritionist, a physical therapist, a bariatric surgeon, and the residency associate programme director who spearheaded the MODS. MODS aligns with the educational curriculum from the Certified Obesity Medical Physician (COMP) programme by the Obesity Society and from the American Board of Bariatric Medicine (ABBM) programme by the American Society of Bariatric Physicians.Box 1 details the specific topics delivered in the MODS curriculum.
(Enlarge Image)
Figure 1.
Multidisciplinary obesity-specific didactic sessions (MODS) curriculum process map.
To evaluate IM residents' knowledge and attitudes towards obesity management, an Obesity Awareness Questionnaire (OAQ) (see online supplementary appendix A http://pmj.bmj.com/content/90/1069/630/suppl/DC1) was administered to the IM residents 2 weeks before the intervention and again 6 months later. The OAQ questions were randomly selected from the validated questionnaire developed by the ABBM and COMP programmes. It consisted of 10 questions on general obesity knowledge and 10 questions on attitudes towards obesity management. Pre- and post-OAQ resident data were collected and reported in aggregate due to regulations by our IRB and residency administration for participant de-identification. The expert lecturers were blinded to the OAQ contents.
To evaluate IM residents' practice behaviours toward obesity management, the following seven guideline-specific performance measures were tracked through retrospective chart reviews of eligible patients managed in the residents' continuity clinics: (1) referrals to nutritionist/dietitian; (2) referrals to other community resources including exercise physical therapist and/or psychiatry for food addiction/anxiety/depression as indicated; (3) referrals to bariatric surgery; (4) initiation of obesity-related medicines (eg, orlistat) or adjustment of medication regimen to avoid weight gain side-effects; (5) obesity listed as an active problem list; (6) development and counselling of dietary plan, physical activity, or behavioural modification plan recorded in the note; (7) frequency of glycated haemoglobin (%HbA1c) or low density lipoprotein (LDL) cholesterol measurement (if concurrent diabetes or hyperlipidaemia, respectively). Time periods compared were the 4 months before and 6 months after the MODS intervention.
The primary patient-specific clinical outcome measure was change in BMI pre- (4 months before) and post- (6 months after) MODS intervention. Secondary clinical outcomes were %HbA1c and LDL cholesterol values (if concurrent diabetes and hyperlipidaemia, respectively). Goals were BMI <25 kg/m, %HbA1c <7%, and LDL cholesterol <100 mg/dL.
This study compared outcomes 4 months before and 6 months after exposure of the resident to the MODS intervention. The main outcomes analysed were resident knowledge and attitudes. The secondary outcomes were patient-specific clinical outcomes and residents' clinical practice behaviours. Data abstraction for patient-specific clinical outcomes and residents' clinical practice behaviours was done retrospectively by three consistent individuals, utilising a standardised chart abstraction checklist to reduce inter-rater variability. For further consistency, one individual was assigned to one specific clinic site for data abstraction and retrospectively reviewed the medical record of every eligible study patient scheduled in each individual resident's clinic at that site. Due to IRB regulation for subject de-identification, pre- and post-intervention patient and resident data were collected and reported in aggregate.
All inferential analyses were paired comparisons pre- versus post-intervention. The one sample t test was utilised for quantitative outcomes. The exact McNemar's test was utilised for binary data. ORs represent conditional ORs given pair is discordant (one positive, one negative). Statistical Analysis Systems (SAS), V.9.3 was utilised for all analyses.
For the OAQ, the total score (number of correct answers) was treated as a continuous variable. Based on experience with similar tests, we expected the total score to follow a Gaussian ('normal') distribution. For patient-resident dyadic binary data, comparisons were made by ORs. Because this study was mainly descriptive, we did not control for study wise error. Significant findings need to be verified by an independent study before considering them definitive.
De-identification of the patients prevented us from conducting subset analysis based on patient-resident dyads.
Methods
Study Setting
This study ran from July 2010 to June 2011 and was approved by the University of Florida (UF) Institutional Review Board (IRB).
The UF IM Residency programme is a large university-based postgraduate programme with 10 medicine subspecialty divisions, training 75 IM residents over a 3-year period (also known as categorical residents). Residents receive formal didactic education on important research and clinical topics in IM through once weekly mortality and morbidity conferences, once weekly medicine grand rounds, daily 1 h morning reports, daily 1 h resident noon conferences, and monthly 1 h conferences before continuity clinic. For experiential training, residents rotate through two large hospitals—Malcolm Randall Veterans Affairs Medical Center (VAMC), the second largest US veterans' hospital system with 239 beds; and UF Health Shands Hospital, an 850-bed, non-profit tertiary care referral centre for the south-eastern USA. Categorical residents are also assigned to a 3-year continuity clinic at one of three sites (UF Medical Plaza, UF Tower Hill, or VAMC) where they provide ambulatory care to a longitudinal panel of primary care patients one half-day per week.
Participants
Physician-participants included all postgraduate year 1 (PGY-1), PGY-2, and PGY-3 categorical IM residents in the 2010–2011 academic year who had managed a panel of their own continuity clinic patients for at least 4 months before and 6 months after the multidisciplinary obesity-specific didactic sessions (MODS) curriculum. Being in a 3-year postgraduate training programme, these residents were present for our entire educational intervention.
Medical records of patients with body mass index (BMI) >25 kg/m in the three IM residents' clinic sites were reviewed retrospectively for clinical outcomes and residents' clinical practice behaviours related to obesity management. The patient population in our large urbanised academic clinic consisted of a 1.5 :1 female : male ratio with average age of 52 years old, a variety of chronic medical conditions (ie, coronary artery disease, chronic obstructive pulmonary disease, diabetes), and a sizeable population of vulnerable, indigent patients (35% non-Caucasian, 13% Medicaid, 10% uninsured). Patient eligibility included age >18 years, BMI >25 kg/m, and established care in the IM residents' clinics at least 4 months before and 6 months after the MODS intervention. The only exclusion criterion was pregnancy.
Educational Intervention: Multidisciplinary Obesity-specific Didactic Sessions
The educational intervention consisted of four 1 h obesity-specific didactic sessions scheduled during the IM resident noon conference every 2 weeks over an 8-week time period from 5 November 2010 to 17 December 2010 (figure 1). These lectures were delivered by a multidisciplinary team of academic expert lecturers with extensive research and clinical training in obesity medicine. The team included an endocrinologist, a psychiatrist, a public health physician, a nutritionist, a physical therapist, a bariatric surgeon, and the residency associate programme director who spearheaded the MODS. MODS aligns with the educational curriculum from the Certified Obesity Medical Physician (COMP) programme by the Obesity Society and from the American Board of Bariatric Medicine (ABBM) programme by the American Society of Bariatric Physicians.Box 1 details the specific topics delivered in the MODS curriculum.
(Enlarge Image)
Figure 1.
Multidisciplinary obesity-specific didactic sessions (MODS) curriculum process map.
Resident Knowledge and Attitudes Measures
To evaluate IM residents' knowledge and attitudes towards obesity management, an Obesity Awareness Questionnaire (OAQ) (see online supplementary appendix A http://pmj.bmj.com/content/90/1069/630/suppl/DC1) was administered to the IM residents 2 weeks before the intervention and again 6 months later. The OAQ questions were randomly selected from the validated questionnaire developed by the ABBM and COMP programmes. It consisted of 10 questions on general obesity knowledge and 10 questions on attitudes towards obesity management. Pre- and post-OAQ resident data were collected and reported in aggregate due to regulations by our IRB and residency administration for participant de-identification. The expert lecturers were blinded to the OAQ contents.
Residents' Clinical Practice Behaviours
To evaluate IM residents' practice behaviours toward obesity management, the following seven guideline-specific performance measures were tracked through retrospective chart reviews of eligible patients managed in the residents' continuity clinics: (1) referrals to nutritionist/dietitian; (2) referrals to other community resources including exercise physical therapist and/or psychiatry for food addiction/anxiety/depression as indicated; (3) referrals to bariatric surgery; (4) initiation of obesity-related medicines (eg, orlistat) or adjustment of medication regimen to avoid weight gain side-effects; (5) obesity listed as an active problem list; (6) development and counselling of dietary plan, physical activity, or behavioural modification plan recorded in the note; (7) frequency of glycated haemoglobin (%HbA1c) or low density lipoprotein (LDL) cholesterol measurement (if concurrent diabetes or hyperlipidaemia, respectively). Time periods compared were the 4 months before and 6 months after the MODS intervention.
Clinical Outcomes Measures
The primary patient-specific clinical outcome measure was change in BMI pre- (4 months before) and post- (6 months after) MODS intervention. Secondary clinical outcomes were %HbA1c and LDL cholesterol values (if concurrent diabetes and hyperlipidaemia, respectively). Goals were BMI <25 kg/m, %HbA1c <7%, and LDL cholesterol <100 mg/dL.
Data Analysis
This study compared outcomes 4 months before and 6 months after exposure of the resident to the MODS intervention. The main outcomes analysed were resident knowledge and attitudes. The secondary outcomes were patient-specific clinical outcomes and residents' clinical practice behaviours. Data abstraction for patient-specific clinical outcomes and residents' clinical practice behaviours was done retrospectively by three consistent individuals, utilising a standardised chart abstraction checklist to reduce inter-rater variability. For further consistency, one individual was assigned to one specific clinic site for data abstraction and retrospectively reviewed the medical record of every eligible study patient scheduled in each individual resident's clinic at that site. Due to IRB regulation for subject de-identification, pre- and post-intervention patient and resident data were collected and reported in aggregate.
Statistical Analysis
All inferential analyses were paired comparisons pre- versus post-intervention. The one sample t test was utilised for quantitative outcomes. The exact McNemar's test was utilised for binary data. ORs represent conditional ORs given pair is discordant (one positive, one negative). Statistical Analysis Systems (SAS), V.9.3 was utilised for all analyses.
For the OAQ, the total score (number of correct answers) was treated as a continuous variable. Based on experience with similar tests, we expected the total score to follow a Gaussian ('normal') distribution. For patient-resident dyadic binary data, comparisons were made by ORs. Because this study was mainly descriptive, we did not control for study wise error. Significant findings need to be verified by an independent study before considering them definitive.
De-identification of the patients prevented us from conducting subset analysis based on patient-resident dyads.
Source...