Volume 19, Issue 2, Fall/Winter 2019

Project Update

Diabetes Infrastructure for Surveillance Evaluation and Research (DISER) Update

We have been running the Alberta Caring for Diabetes (ABCD) cohort study for a number of years now. We initially recruited over 2,000 adults living with type 2 diabetes, and have asked these individuals to complete surveys on a regular basis over the past 5-6 years. Most recently, we implemented an on-line survey option, rather than traditional paper and pencil surveys sent in via snail mail. The ABCD cohort has provided data for a number of ACHORD projects, which have been used for student thesis research, and for informing activities in the province, such as understanding foot self-care activities for people with type 2 diabetes.

One of the original objectives of the ABCD cohort study was to link self-reported self-care management data collected to administrative databases that are used for population-based surveillance activities. Through this linkage, it would be possible to explore relationships between self-care behaviours and health care utilization and health outcomes. As a partner in the Diabetes Infrastructure for Surveillance, Evaluation and Research (DISER) project, we are working with the Diabetes, Obesity and Nutrition SCN from Alberta Health Services (AHS) to update the linkages with data in the Clinical Analytics repositories.

Of the original 2040 ABCD cohort participants, 1871 consented to have their survey data linked with such administrative data. For these individuals, data has been retrieved from AHS databases, including the following elements:
* Physician Claims containing information on claims submitted for payment of Alberta service providers, and information on location of provider, patient demographics, practitioner type, what the service was claimed for, date of service and associated diagnosis code (coded in ICD-9-CA)
* Discharge Abstract Database (DAD) - continuing information on discharges from all acute care institutions in the province
* Alberta Ambulatory Care Reporting System (AACRS), and National Ambulatory Care Reporting System (NACRS) - containing information on visits to ambulatory care institutions across the province such as emergency departments, urgent care and day surgery
* Pharmaceutical Information Network (PIN) – containing all information on prescribing drug dispenses from community-based pharmacies across the province, including patient demographics, drug name, drug identification number, WHO ATC code and dispense date; and
* Laboratory Results Data – containing information on all general lab tests performed across the province sourced from inpatient and outpatient lab, such as HbA1c, fasting blood glucose tests and random blood glucose tests.

These linked data are available from the time the first physician claim was made with a diagnosis code for diabetes and an infinite look back at all medications, laboratory tests (HbA1c, blood glucose), hospitalizations, and emergency room visits associated with diabetes.

In one of our first DISER projects using these linked data, together with the Clinical Analytics group of the DON SCN, we are aiming to test the utility of a new case definition to identify diabetes, compared with the widely used National Diabetes Surveillance System (NDSS) case definition. Specifically we will use the NDSS case definition to identify individuals with diabetes compared with a case definitions using PIN and Laboratory data. To enhance the comparison of these data, we will use the ABCD cohort (a known cohort of people living with diabetes) to assess the sensitivity of these case definitions for identifying individuals with diabetes.

We welcomed Sabrina Saba, a new masters student, in September whose work will use these DISER data to examine relationships between self-care dietary behaviours and health care utilization.

The DISER data are available for use for other projects. Please contact me nonsi.mathe@ualberta.ca or call 780-492-9257 to discuss possible work.

Recent Literature from the ACHORD Journal Club

(Paper discussed Wednesday, October 9, 2019; Commentary by Jiabi Wen)

Versteegh MM, Ramos IC, Buyukkaramikli NC, Anasaripour A, Reckers-Droog VT, Brouwer WBF. Severity-Adjusted Probability of Being Cost Effective. Pharmacoeconomics (2019) 37(9):1155-1163. PMID:31134467.

Background
Cost-effectiveness analysis compares incremental costs and incremental effectiveness of different treatments in terms of cost-effectiveness ratio, i.e. costs per quality-adjusted life years (QALYs). This ratio indicates the additional money to spend for every additional one QALY. If a cost-effectiveness ratio gets a value below the incremental cost-effectiveness ratio (ICER), an external criterion, it indicates a specific treatment has better value for money versus its comparator. In the context of priority setting, some countries adopt differential ICERs to reflect their higher willingness to pay under some situation, e.g. health gains for the worse-off. For example, the Netherlands defines three equity classes based on disease severity, and the highest threshold, €80,000 per QALY, applies to the most severe condition. But there is uncertainty in the measurement of severity, leading to uncertainty in the choose of a threshold. This paper explored a new method, using the severity-adjusted probability of being cost-effective, to better inform decision-making.

Article Summary:
In this paper, the authors first standardized the assessment of severity. It is defined by the shortfall of quality-adjusted life expectancy (QALE) between the patient group and the non-ill (i.e., general) population (age and gender controlled). Both QALEs consider a lifetime horizon, but do not include discounting. Normal QALE is the expected total number of undiscounted QALYs the non-ill controlled general population attains, and here derived from population norms. Patient QALE is equal to the number of QALYs for the population of interest based on standard of care. In this regard, the more effective the current treatment is, the smaller the gap exists in current situation.

The uncertainty behind point estimates of severity is a common issue. The uncertainty mainly comes from the heterogeneity in quality-of-life estimates, i.e. the variation in health states and the public’s preference for those health states. The uncertainty may lead to a different decision in categorizing severity class, and therefore the different ICER threshold to apply leading to uncertainty in the cost-effectiveness results. To address this issue, the authors combine the uncertainty of severity with the uncertainty in a cost-effectiveness model. For the uncertainty in the cost-effectiveness model, Monte-Carlo simulation is a commonly-used method. By drawing k-replications from the distributions of different parameters, k-pairs of QALY gains (∆Q_i) and costs (∆C_i) are generated. The probability that equation v-threshold×∆Q_i-∆C_i>0 holds is the probability of the new technology being cost-effective, where v-threshold is the specific ICER threshold based on the severity-adjusted criteria. In this regard, integrating the severity uncertainty into probability sensitivity analysis is varying the v-threshold based on the severity class in each round. k-replications of severity estimates are drawn, and each estimate determines a level of ICER. The probability that the equation 〖v-threshold〗_i×∆Q_i-∆C_i>0 holds is the severity-adjusted probability of the new technology being cost effective.

The authors did a case study to show the difference between using point estimates of severity and considering the uncertainty of severity. In their hypothetical oncology treatment, the point estimate of disease severity indicates an ICER of €80,000, but the confidence interval of the disease severity crosses two equity class, which shows that both €50,000 and €80,000 are potentially applicable. The probability of being cost-effective under two thresholds are 3% and 85%, respectively. The severity-adjusted probability of being cost-effective is 62.4%, which may reduce the chance of an incorrect decision due to the technology being evaluated against the ‘wrong’ threshold.

Discussion Within Journal Club:
Comparing with the categorical differential thresholds, the severity-adjusted probability of being cost-effective better reflects the uncertainty and heterogeneity in the calculation. But there are some drawbacks and we discussed these in the journal club. Firstly, the estimate of normal QALE has a systematic bias. The comparator population includes the patient for which severity is being calculated as the population norms are derived based on the whole (i.e., general) population. Only if the disease is relatively rare, the bias could be neglected. We thought this could be solved by establishing the quality-of-life database with a filter on common conditions. Secondly, although this might be a helpful method to avoid the ‘wrong’ decision in applying a threshold, the rationale behind using different thresholds for each severity class remains open to debate. We further talked about technology efficiency and allocative efficiency, and the difference between demand side (i.e., willingness to pay), and supply side threshold. A cost-effective result simply means that the technology is better than its comparator at some specified threshold. It may not indicate the worthwhileness of allocating more budget to fund a treatment for this condition rather than other health conditions (i.e., opportunity cost). For example, allocating more budget to orphan drugs may create shortages in the budget for other conditions. Orphan drugs can be expensive, and the total number of people living with rare diseases is substantial (e.g. 7.5%-9% in America). Regardless of a specific treatment’s cost-effectiveness, equity issues are always important to consider. Journal club members agreed that methods of measuring allocative efficiency should be proposed in the future.

ACHORD Seen and Heard

Recent Publications

Campbell-Scherer DL, Asselin J, Osunlana AM, Ogunleye AA, Fielding S, Anderson R, Cave A, Johnson JA, Sharma AM. Changing provider behaviour to increase nurse visits for obesity in family practice: The 5As Team Randomised Control Trial. CMAJ Open 2019;7(2):e371-e378. PMID: 31147378.

Eskin M, Simpson SH, Eurich DT. Evaluation of Health User Effects with Metformin and Other Oral Antihyperglycemia Medication Users in Adult Patients with Type 2 Diabetes. Can J Diabetes 2019;43(5):322-328. PMID: 30782471.

Al Sayah F, Yeung R, Johnson JA. Association of Depressive Symptoms and Diabetes Distress with Severe Hypoglycemia in Adults with Type 2 Diabetes. Can J Diabetes 2019;43(5):316-321. PMID: 30578165.

Majumdar SR, Lier DA, McAlister FA, Johnson JA, Rowe BH, Beaupre LA. Cost-effectiveness of Osteoporosis Interventions to Improve Quality of Care after Upper Extremity Fracture: Results from a Randomized Trial (C-STOP Trial). J Bone Miner Res (JBMR). 2019;34(7):1220-1228. PMID: 30779861.

Ye M, Vena JE, Johnson JA, Xu JY, Eurich DT. Validation of Drug Prescription Records for Senior Patients in Alberta’s Tomorrow Project: Assessing agreement between two population-level administrative pharmaceutical databases in Alberta, Canada. Pharmacoepidemiology & Drug Safety 2019;28(10):1417-1421. PMID: 31348593.

Ye M, Robson PJ, Eurich DT, Vena JE, Xu JY, Johnson JA. Anthropometric Changes and Risk of Diabetes: Are There Differences between Men and Women? A Longitudinal Study of Alberta’s Tomorrow Project Cohort. BMJ Open 2019;9(7):e023829. PMID: 31326923.

Zongo A, Simpson SH, Johnson JA, Eurich DT. Optimal threshold of adherence to lipid-lowering drugs in predicting acute coronary syndrome, stroke, or mortality: a cohort study. PLOS One 2019;14(9):e0223062. PMID: 31553787.

Burton CE, Sester M, Robinson JL, Eurich DT, Urschel S, Preiksaitis JK. CMV-Specific T-cells and CD27-CD28-CD4+ T-cells for Assignment of Cytomegalovirus (CMV) Status in Adults Awaiting Organ Transplant. J Clin Virol. 2019;115:37-42 PMID: 30959325.

Aubin D, Hebert M. Eurich DT. The importance of measuring the impact of patient-oriented research. CMAJ. 2019;191(31):E860-E864. PMID: 31387956.

Sharma V, Weir D, Samanani S, Simpson SH, Gilani F, Jess E, Eurich DT. Characterisation of concurrent use of prescription opioids and benzodiazepine/Z-drugs in Alberta, Canada: a population-based study. BMJ Open. 2019;9(9):e030858. PMID: 31494618.

Hopkins JJ, Reif R, Bigam D, Baracos VE, Eurich DT, Sawyer MM. Change in Skeletal Muscle Following Resection of Stage I–III Colorectal Cancer is Predictive of Poor Survival: A Cohort Study. World J Surg. 2019;43(10):2518-2526. PMID: 31222643.

Meeting Presentations and Invited Talks

Wen J, Jin X, Ohinmaa A. Economic Evaluation of Sucrose Octasulfate Dressing for the Treatment of Diabetic Foot Ulcer in Canada. 11th Huaxia Pharmacoeconomics and HTA Forum, Nanjing, China, June 29-30, 2019.

Eurich DT, Smith S, Ye M, Hui DSC, Zelyas N, Zapernick L, Wong R, Labib M. Chan PKS, Lee N. Development Of An Ordinal Scale Treatment Endpoint In Adults Hospitalized With Influenza. Options X, for the Control of Influenza. Suntec City, Singapore, August 28, 2019.

Ye M, Eurich D, Vena J, Johnson J. Linking Administrative healthcare data to Alberta’s Tomorrow Project: Building A Better Platform for Cancer and Chronic Disease Research. Alberta's Tomorrow Project Scientific Steering Committee Meeting, Calgary, Alberta, Canada, September 26, 2019.

Short H, Al Sayah F, Ohinmaa A, Johnson JA. Screening for anxiety and depressive symptoms using the EQ-5D-3L post-hospital discharge. ISOQOL 26th Annual Conference, San Diego, California, United States, October 20-23, 2019.

Soprovich AL, Seaton CL, Bottorff JL, Duncan MJ, Caperchione CM, Oliffe JL, James C, Rice S, Tjosvold L, Eurich DT, Johnson ST. REST Up: A systematic review and meta-analysis of workplace behavioural interventions to promote sleep health in men. CMHA Working Stronger, Edmonton, Alberta, Canada, October 28-29, 2019,

Short H, Al Sayah F, Ohinmaa A, Johnson JA. Screening for anxiety and depressive symptoms using the EQ-5D-3L post-hospital discharge. This is Public Health Week, Edmonton, Alberta, Canada, November 4-8, 2019.

Soprovich A and Al Sayah F. Patient-Reported Outcomes (PROs) in Primary Care Research: An introduction to various applications and analyses. 47th North American Primary Care Research Group (NAPCRG) Annual Meeting, Toronto, Ontario, Canada, November 16-20, 2019.

ACHORD
University of Alberta
2-040 Li Ka Shing CHRI
Edmonton, AB T6G 2E1
email: achord@ualberta.ca
Phone: 780-248-1010