Quality gaps in care delivery among emerging adults with type 1 diabetes: A retrospective cohort study

Emerging adulthood (EA), between 18 and 30 years of age, represents a high-risk period for patients with type 1 diabetes mellitus (T1DM).1-4 Loss-to-follow-up rates (no appointment for >12 months) range from 25% to 40% up to years after discharge from pediatrics among T1DM EA patients,3, 5 which can compromise patient education and complication screening, and result in worsened glycemic control6, 7 and increased hospitalizations rates.8 This study aims to understand transition care quality gaps in an interdisciplinary T1DM EA clinic in a large Canadian city. This study was undertaken in an interdisciplinary T1DM EA (18–25 years) program at an academic, ambulatory hospital in Toronto, Canada, consisting of four endocrinologists, two nurses, one dietician, and one social worker. A baseline audit of nonattendance rates of new and follow-up T1DM EA was conducted between February 1, 2015 and September 30, 2016, from the hospital's electronic medical record system (Epic©). Nonattendance was defined as all missed scheduled appointments (no-shows and cancellations <24 h) plotted monthly on a statistical P-chart. A retrospective chart review was conducted on all consecutive new patients referred to the EA program from February 1, 2015 to September 30, 2015. Information was collected for 1) demographic information, 2) attendance data, 3) glycemic/metabolic control, 4) routine screening (per Diabetes Canada Clinical Practice Guidelines [CPG]9), 5) acute diabetes complications, and 6) counseling documentation.9 Descriptive analyses were utilized with continuous variables reported as medians (interquartile range [IQR]) and categorical variables reported as percentages. χ2 testing was used to compare categorical outcomes and Mann–Whitney U test to compare medians assessed for significance (p < 0.05). A total of 444 missed encounters were registered for 150 patients over 20 months. Mean nonattendance rate was 31.2% (Figure 1) with the no-show rate (missed appointment without documented cancellation) at 23.7%. Statistical process control P-chart depicting the baseline nonattendance rate of emerging adult patients at the Young Adult Diabetes Program. All prebooked in-person clinical encounters were documented and the proportion of missed appointments to attended appointments was calculated to determine a nonattendance ratio per monthly basis. Missed appointments included both no-show appointments and cancellations within less than 24 h. A total of 444 missed clinical encounters were registered for 150 patients over a 20-month period. Upper control limit (UCL) = 0.694; center line (CL) = 0.312; and lower control limit (LCL) = 0.000. y-axis = nonattendance rate (as a proportion of all in-person clinical encounters); x-axis = time (in months). Fifty-one new referrals were registered between February 1, 2015 and September 30, 2015. The median age at the first visit was 22.9 years (IQR: 2.9 years) with 84.3% identifying female, and a median initial %HbA1c of 8.5% (IQR: 1.9%). Fourteen patients (27.5%) had comorbid autoimmune conditions and 11 (21.5%) had a pre-existing psychiatric diagnosis. Nine patients (17.6%) had a previous severe hypoglycemic event and six (11.7%) had a previous diabetes-associated hospitalization. The baseline attendance rate of 51 new patients is presented in Table 1A. There were 13.7% of patients lost-to-follow-up. Of those with follow-up, the median number of visits per year per patient was 3.7, with a median of 2.7 visits with the endocrinologist. Overall, 35.3% of patients missed ≥2 appointments. Of those with ≥2 missed visits (nonattenders) compared to <2 missed visits (attenders), nonattenders had significantly lower median frequency of self-blood-glucose monitoring checks compared to attenders (2 vs. 3, respectively, p < 0.05), lower insulin pump use (45.5% vs. 5.6%, respectively, p < 0.05), and higher prior hospitalization rates for diabetes-related complication (22.2% vs. 3%, respectively, p < 0.05) (Table 1B). There was no difference in age at the first visit, diabetes duration, %HbA1c, micro/macrovascular complications, and prior severe documented hypoglycemic events between nonattenders and attenders. ≥2 Missed visit(s)a (N = 18) Nonattenders Table 1C illustrates the frequency of CPG screening and counseling documented.9 Patients were consistent with regular %HbA1c checks at 6-month intervals and annual eye screening (100% and 97%, respectively). Annual monofilament testing (69%) and lipid profile screening (67%) were inconsistent. Smoking status was documented 61% of the time. The frequency of documented counseling across 11 topics was variable. Preconception counseling, driving and hypoglycemia counseling, and hypoglycemia management counseling were most frequently discussed at 79.1% (N = 43, female patients), 70.6% (N = 51), and 68.6% (N = 51), respectively. The most infrequently documented counseling topics were medical alert counseling at 27.5% (N = 51), eating disorder screening at 21.6% (N = 51), and sick-day management at 7.8% (N = 51). We noted a nonattendance rate of 31%. Despite being structurally like other EA T1DM programs, our nonattendance rate was similar to general diabetes clinics and higher than the reported 12%–16% rate of other T1DM EA programs.10-13 This may be due to cancellations <24 h being excluded from other nonattendance definitions. Higher nonattendance rates may result from a lack of a resource-intensive transition coordinator.11, 12, 14, 15 Joint pediatric-adult diabetes appointments have been effective for attendance,12, 13 but not feasible in Ontario where the pediatric cutoff age is provincially mandated. Our interdisciplinary team was only available during working hours, but some studies have benefitted from extended clinic hours for scheduling flexibility.11, 14 Although our EA program uses phone call reminders, email/text reminders were not permitted, which have been effective in improving attendance.10, 12, 14, 15 This suggests the need for context and resource-specific interventions, and the need to address organizational barriers to streamline team communication. Nonattenders had infrequent self-blood-glucose monitoring, were less likely to be insulin pump users, and had higher rates of diabetes-associated hospitalizations compared to attendees. Nonattenders trended towards higher baseline %HbA1c values, although this was not significant. Similarly, in the literature, predictors of nonattendance include multiple-daily injections, higher %HbA1c levels, and fewer physician visits before adult care.15 This suggests a need to identify high-risk patients early and optimize diabetes management pretransition. Patients with closer monitoring and fewer complications were less likely to miss appointments in adult care. Our study corroborates similar findings on variability in diabetes complication screening as a quality gap in transition care. An Australian study in a T1DM EA cohort16 showed only 12%–14% of patients having a documented ophthalmic examination, and 30.8%–32.6% had a documented albumin-to-creatinine ratio measurement over a 2-year period.16 In our study, only 61% had a documented smoking screening assessment. One study reported only 30.4% of youth aged 10–14 years were asked about smoking, while only 47.2% were counseled on smoking cessation.17 Although screening guidelines have proven effective at informing clinical decision-making, there continue to be application gaps. Nonattendance among T1DM EA may contribute to inconsistency in guideline application. Overall, 21.6% of patients had documentation of eating disorder assessment in our study, indicating a need for more systematic screening of disordered eating; up to 10% of T1DM have an eating disorder.18 In our study, 79.1% of female patients received preconception counseling; this is important as a study demonstrated that knowledge of outcomes of uncontrolled diabetes and fetal development is unsatisfactory, but patient interest in receiving competent preconception education by a diabetologist is high (88.6%).19 Our study quantifies the rate of documented mood screening at 64.7%. A qualitative study of an Australian T1DM cohort (18–30 years) demonstrated little patient understanding of increased mental health risks.20 Our study reported documented sick day counseling, medical alert bracelet counseling, patient goal setting, and alcohol counseling rates at 7.8%, 27.5%, 58.8%, and 62.7%, respectively. To date, the frequency of these screening practices had not been reported. Limitations in our study include a predominantly female and small patient cohort. Furthermore, our chart review represented a cohort that was receiving regular diabetes follow-up; as such, there is underrepresentation from the population of patients who are loss-to-follow. We were unable to acquire data on diabetes-related admissions or emergency room attendances between attenders and non-attenders. This study highlights important quality gaps in transition care delivery for T1DM EA. Future studies involving the implementation of patient-centered interventions, while being mindful of local contextual factors and resources, will be an important way of improving care delivery of T1DM EA. Geetha Mukerji: Conceptualization, resources, and supervision. Xinye S. Wang: Data curation and investigation. Xinye S. Wang and Husayn Marani: Formal analysis and visualization. Geetha Mukerji and Xinye S. Wang: Methodology. Xinye S. Wang, Geetha Mukerji, and Husayn Marani: Writing–original draft preparation. Cheryl Harris-Taylor, Leah Drazek, Janis Rusen, Nicola Farnell, and Lorraine Lipscombe: Writing–review and editing. All authors have read and approved the final version of the manuscript. The authors would like to thank Dr. Jenny Y. Wang for her role in laying the foundation for this study and for her contributions to the design, analysis, and execution of the project. They would also like to thank the patients living with type 1 diabetes within the EA Diabetes Program, Patrizia Diraimo, Kishani Sarvananthan, Victor Cipriano, Eleni Dimaraki, Arno Kumagai, and the Women's College Hospital Institute for Health System Solutions and Virtual Care. Lorraine Lipscombe acknowledges the funding support by a Diabetes Investigator Award from Diabetes Canada. The authors declare no conflict of interest. The study protocol was reviewed and approved by the Women's College Hospital Ethics Assessment Process for Quality Improvement Projects. The lead author Geetha Mukerji affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained. Data are available upon request by contacting the corresponding author, Dr. Geetha Mukerji. Dr. Geetha Mukerji had full access to all of the data in this study and takes complete responsibility for the integrity of the data and the accuracy of the data analysis.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.