Optum Labs: Building A Novel Node In The Learning Health Care System

HbA1c overtesting and overtreatment among US adults with controlled type 2 diabetes, 2001-13: observational population based study

STUDY QUESTION

What is the extent and effect of excessive testing for glycated hemoglobin (HbA1c) among adults with controlled type 2 diabetes?

METHODS

A retrospective analysis of data from a national administrative claims database included commercially insured individuals in the USA, 2001-13. Study patients were aged 18 years or older, had type 2 diabetes with stable glycemic control (two consecutive tests showing HbA1c<7.0% within 24 months), did not use insulin, had no history of severe hypoglycemia or hyperglycemia, and were not pregnant. HbA1c testing frequency was measured within 24 months after the second (index) HbA1c test, and classified as guideline recommended (≤ 2 times/year), frequent (3-4 times/year), and excessive (≥ 5 times/year). Changes in treatment regimen were ascertained within three months of the index test.

STUDY ANSWER AND LIMITATIONS

Of 31,545 patients in the study cohort (mean age 58 years; mean index HbA1c 6.2%), HbA1c testing frequency was excessive in 6% and frequent in 55%. Despite good glycemic control at baseline, treatment was further intensified by addition of glucose lowering drugs or insulin in 8.4% of patients (comprising 13%, 9%, and 7% of those tested excessively, frequently, and per guidelines, respectively; P<0.001). Compared with guideline recommended testing, excessive testing was associated with treatment intensification (odds ratio 1.35 (95% confidence interval 1.22 to 1.50)). Excessive testing rates remained unchanged in 2001-08, but fell significantly after 2009. The odds of excessive testing was 46% lower in 2011 than in 2001-02. The study population is not representative of all US patients with type 2 diabetes because it was restricted to commercially insured adults with stable and controlled diabetes not receiving insulin treatment. The study design did not capture the underuse of HbA1c testing.

WHAT THIS STUDY ADDS

In this US cohort of adults with stable and controlled type 2 diabetes, more than 60% received too many HbA1c tests, a practice associated with potential overtreatment with hypoglycemic drugs. Excessive testing contributes to the growing problem of waste in healthcare and increased patient burden in diabetes management.

FUNDING, COMPETING INTERESTS, DATA SHARING

NDS and RGM are funded partly by the Agency for Healthcare Research and Quality (R18HS18339) and AcademyHealth Delivery System Science Fellowship (2013), respectively. No competing interests declared. Additional data are available from mccoy.rozalina@mayo.edu.

Advancing the Use of Administrative Data for Emergency Department Diagnostic Imaging Research

Administrative data are critical to describing patterns of use, cost, and appropriateness of imaging in emergency care. These data encompass a range of source materials that have been collected primarily for a nonresearch use: documenting clinical care (e.g., medical records), administering care (e.g., picture archiving and communication systems), or financial transactions (e.g., insurance claims). These data have served as the foundation for large, descriptive studies that have documented the rise and expanded role of diagnostic imaging in the emergency department (ED). This article summarizes the discussions of the breakout session on the use of administrative data for emergency imaging research at the May 2015 Academic Emergency Medicine consensus conference, "Diagnostic Imaging in the Emergency Department: A Research Agenda to Optimize Utilization." The authors describe the areas where administrative data have been applied to research evaluating the use of diagnostic imaging in the ED, the common sources for these data, and the strengths and limitations of administrative data. Next, the future role of administrative data is examined for answering key research questions in an evolving health system increasingly focused on measuring appropriateness, ensuring quality, and improving value for health spending. This article specifically focuses on four thematic areas: data quality, appropriateness and value, special populations, and policy interventions.

Pacemaker implantation after catheter ablation for atrial fibrillation

BACKGROUND

Sinus node dysfunction requiring pacemaker implantation is commonly associated with atrial fibrillation (AF), but may not be clinically apparent until restoration of sinus rhythm with ablation or cardioversion. We sought to determine frequency, time course, and predictors for pacemaker implantation after catheter ablation, and to compare the overall rates to a matched cardioversion cohort.

METHODS AND RESULTS

We conducted a retrospective analysis using a large US commercial insurance database and identified 12,158 AF patients who underwent catheter ablation between January 1, 2005 and December 31, 2012. Over an average of 2.4 years of follow-up, 5.6 % of the patients underwent pacemaker implantation. Using the Cox proportional hazards models, we found that risk of risks of pacemaker implantation was associated with older age (50-64 and ≥65 versus <50 years), female gender, higher CHADS2 score (≥2 and 1 versus 0), higher Charlson index (≥2 versus 0-1), certain baseline comorbidities (conduction disorder, coronary atherosclerosis, and congestive heart failure), and the year of ablation. There was no significant difference in the risk of pacemaker implantation between ablation patients and propensity score (PS)-matched cardioversion groups (3.5 versus. 4.1 % at 1 year and 8.8 versus 8.3 % at 5 years).

CONCLUSION

Overall, pacemaker implantation occurs in about 1/28 patients within 1 year of catheter ablation. The overall implantation rate decreased between 2005 and 2012. Furthermore, the risk after ablation is similar to cardioversion, suggesting that patients require pacing due to a common underlying electrophysiologic substrate, rather than the ablation itself.

Patterns of Anticoagulation Use and Cardioembolic Risk After Catheter Ablation for Atrial Fibrillation

Background

There is significant practice variation in oral anticoagulation (OAC) use following catheter ablation for atrial fibrillation. It is not clear whether the risk of cardioembolism increases after discontinuation of OAC following catheter ablation.

Methods and Results

We identified 6886 patients within a large national administrative claims database who underwent catheter ablation for atrial fibrillation between January 1, 2005, and September 30, 2014. We assessed the effect of time off of OAC by CHA₂DS₂‐VASc score (after adjusting for other comorbidities) on risk of cardioembolism, using Cox proportional hazards models. There was an increase in the use of non–vitamin K OAC after ablation from 0% in 2005 to 69.8% in 2014. OAC discontinuation was high, with only 60.5% and 31.3% of patients remaining on OAC at 3 and 12 months, respectively. The rate of discontinuation was higher in low‐risk patients (82% versus 62.5% at 12 months for CHA₂DS₂‐VASc 0–1 versus ≥2, respectively; P<0.001). Stroke occurred in 1.4% of patients with CHA₂DS₂‐VASc ≥2 and 0.3% of those with CHA₂DS₂‐VASc 0 or 1 over the study follow‐up. The risk of cardioembolism in the first 3 months after ablation was increased among those with any time off OAC (hazard ratio 8.06 [95% CI 1.53–42.3], P<0.05). The risk of cardioembolism beyond 3 months was increased with OAC discontinuation among high‐risk patients (hazard ratio 2.48 [95% CI 1.11–5.52], P<0.05) but not low‐risk patients.

Conclusion

The overall risk of stroke in postablation patients is low; however, OAC discontinuation after ablation is common and is associated with increased risk of cardioembolism for all patients within the first 3 months and for high‐risk patients in the long term. Continuing OAC for at least 3 months in all patients and indefinitely in high‐risk patients appears to be the safest strategy.

Are Patients With Cancer Less Willing to Share Their Health Information? Privacy, Sensitivity, and Social Purpose

PURPOSE

Growing use of electronic health information increases opportunities to build population cancer databases for research and care delivery. Understanding patient views on reuse of health information is essential to shape privacy policies and build trust in these initiatives.

METHODS

We randomly assigned nationally representative participants (N = 3,336) with and without prior cancer to six of 18 scenarios describing different uses of electronic health information. The scenarios varied the user, use, and sensitivity of the information. Participants rated each scenario on a scale of 1 to 10 assessing their willingness to share their electronic health information. We used conjoint analysis to measure the relative importance of each attribute (ie, use, user, and sensitivity).

RESULTS

Participants with and without a prior diagnosis of cancer had a similar willingness to share health information (0.27; P = .42). Both cancer and noncancer participants rated the purpose of information use as the most important factor (importance weights, 67.1% and 45.6%, respectively). For cancer participants, the sensitivity of the information was more important (importance weights, 29.8% v 1.2%). However, cancer participants were more willing to share their health information when the information included more sensitive genetic information (0.48; P = .015). Cancer and noncancer respondents rated uses and users similarly.

CONCLUSION

The information sharing preferences of participants with and without a prior diagnosis of cancer were driven mainly by the purpose of information reuse. Although conventional thinking suggests patients with cancer might be less willing to share their health information, we found participants with cancer were more willing to share their inherited genetic information.

Trends and predictors of readmission after catheter ablation for atrial fibrillation, 2009-2013

As the number of patients undergoing catheter ablation for atrial fibrillation (AF) increases, there is a growing focus on optimizing the quality and efficiency of. Readmission is often considered an indicator of both quality and efficiency of care delivery. We sought to estimate rates and identify predictors of readmission after catheter ablation.

METHODS AND RESULTS

Using a large, national administrative claims database, we identified all AF patients who underwent catheter ablation between January 2009 and December 2013 (10,705 ablation cases). We examined incident readmission and the primary diagnosis during the readmission episode of care. We used Cox proportional hazard models to identify associations between readmission and patient and institutional characteristics. A total of 1,433 (13.4%) ablation patients were readmitted within 90 days of ablation for any cause, and 573 (5.4%) were admitted with AF as the primary diagnosis. There was a decline in all-cause (from 15.6% to 12.8%; P = .04) and AF-related (6.4%-5.0 %; P = .03) 90-day readmission over the study period. In a multivariate model, earlier year of ablation and each of 9 chronic conditions (alone or in combination) were independently associated with risk of readmission.

CONCLUSIONS

Between 2009 and 2013, there was a reduction in 90-day readmission rates after AF ablation, suggesting improved periprocedural care of these patients. Identifying patients at high risk for readmission after catheter ablation for AF may offer an opportunity for early intervention and, ultimately, reduction in procedural morbidity and medical costs.

Effect of Antiarrhythmic Drug Initiation on Readmission After Catheter Ablation for Atrial Fibrillation

OBJECTIVES

This study sought to evaluate the impact on antiarrhythmic drug (AAD) initiation on the risk of readmission after catheter ablation for atrial fibrillation (AF) among patients not already treated with an AAD.

BACKGROUND

Hospital readmission, a commonly tracked indicator of quality and efficiency of care delivery, occurs in about 15% patients within 90 days of undergoing catheter ablation for AF.

METHODS

Using a large national administrative claims database, we identified all atrial fibrillation patients (≥18 years of age) who underwent catheter ablation between January 2005 and December 2013 (n = 7,442). We identified the subset of patients who had not been on an AAD in the 90 days before ablation (n = 2,542) and, among those, the patients in whom an AAD was initiated at discharge following the ablation (n = 519).

RESULTS

The readmission rate was significantly lower among patients who were initiated on an AAD compared with those who were not (11.6% vs. 16.2%, p = 0.009). The association persisted after adjustment for age, sex, Charlson index, and CHADS₂ score (hazard ratio [HR]: 0.73, 95% confidence interval [CI]: 0.56 to 0.97; p = 0.03). In unadjusted time to event analysis, amiodarone (HR: 0.55, 95% CI: 0.32 to 0.94; p = 0.039) was associated with the greatest reduction in readmission whereas dronedarone, Class II agents, and Class IC agents had no statistically significant effect on readmission. AADs were discontinued in 44.5% of patients at 3 months.

CONCLUSIONS

Initiation of an AAD at discharge of catheter ablation is associated with a significant reduction in readmission within 90 days. Routine initiation of an AAD after catheter ablation may reduce healthcare utilization in the periablation period; however, the high rate of medication discontinuation may suggest that side effects or inefficacy may limit long-term AAD use post-ablation.

Use of Post-acute Care Services and Readmissions After Left Ventricular Assist Device Implantation in Privately Insured Patients

BACKGROUND

Very little is known about health care resource utilization, including post-acute care use and hospital readmissions, after left ventricular assist device (LVAD) implantation.

METHODS AND RESULTS

Administrative claims from a database of multiple United States health plans were used to identify patients that received an LVAD (ICD-9 code 37.66) and survived to hospital discharge from January 1-2006, through September 30-2013. Post-acute care use was defined as a skilled nursing facility or rehabilitation stay within 90 days after hospital discharge. Patients were censored at heart transplantation or end of coverage through December 31-2013. Of 583 patients (mean age 55 years, 77% male), 223 (38.3%) used post-acute care services, more commonly in patients with diabetes, who required hemodialysis, and who had LVADs implanted at hospitals in more populated areas, with more beds, and in the northeast region (P < .05 for each). The most common reasons for readmission were device complications, heart failure, and arrhythmia. Readmission risk was higher in patients who had diabetes, peripheral vascular disease, and longer hospital length of stay, but it did not differ by post-acute care use.

CONCLUSIONS

Use of post-acute care services varies based on hospital characteristics. We found no association between post-acute care use and readmission risk after LVAD implantation.

Comparative risk of gastrointestinal bleeding with dabigatran, rivaroxaban, and warfarin: population based cohort study

OBJECTIVE

To determine the real world risk of gastrointestinal bleeding associated with the use of the novel oral anticoagulants dabigatran and rivaroxaban compared with warfarin.

DESIGN

Retrospective, propensity matched cohort study.

SETTING

Optum Labs Data Warehouse, a large database including administrative claims data on privately insured and Medicare Advantage enrollees.

PARTICIPANTS

New users of dabigatran, rivaroxaban, and warfarin from 1 November 2010 to 30 September 2013.

MAIN OUTCOME MEASURES

Incidence rates (events/100 patient years) and propensity score matched Cox proportional hazards models were used to estimate rates of total gastrointestinal bleeds, upper gastrointestinal bleeds, and lower gastrointestinal bleeds for the novel oral anticoagulants compared with warfarin in patients with and without atrial fibrillation. Heterogeneity of treatment effect related to age was examined using a marginal effects model.

RESULTS

The incidence of gastrointestinal bleeding associated with dabigatran was 2.29 (95% confidence interval 1.88 to 2.79) per 100 patient years and that associated with warfarin was 2.87 (2.41 to 3.41) per 100 patient years in patients with atrial fibrillation. In non-atrial fibrillation patients, the incidence of gastrointestinal bleeding was 4.10 (2.47 to 6.80) per 100 patient years with dabigatran and 3.71 (2.16 to 6.40) per 100 patient years with warfarin. With rivaroxaban, 2.84 (2.30 to 3.52) gastrointestinal bleeding events per 100 patient years occurred in atrial fibrillation patients (warfarin 3.06 (2.49 to 3.77)/100 patient years) and 1.66 (1.23 to 2.24) per 100 patient years in non-atrial fibrillation patients (warfarin 1.57 (1.25 to 1.99)/100 patient years). In propensity score matched models, the risk of gastrointestinal bleeding with novel oral anticoagulants was similar to that with warfarin in atrial fibrillation patients (dabigatran v warfarin, hazard ratio 0.79 (0.61 to 1.03); rivaroxaban v warfarin, 0.93 (0.69 to 1.25)) and in non-AF patients (dabigatran v warfarin, hazard ratio 1.14 (0.54 to 2.39); rivaroxaban v warfarin, 0.89 (0.60 to 1.32)). The risk of gastrointestinal bleeding increased after age 65, such that by age 76 the risk exceeded that with warfarin among atrial fibrillation patients taking dabigatran (hazard ratio 2.49 (1.61 to 3.83)) and patients with and without atrial fibrillation taking rivaroxaban (2.91 (1.65 to 4.81) and 4.58 (2.40 to 8.72), respectively).

CONCLUSIONS

The risk of gastrointestinal bleeding related to novel oral anticoagulants was similar to that for warfarin. Caution should be used when prescribing novel oral anticoagulants to older people, particularly those over 75 years of age.

Selecting a dynamic simulation modeling method for health care delivery research-part 2: report of the ISPOR Dynamic Simulation Modeling Emerging Good Practices Task Force

In a previous report, the ISPOR Task Force on Dynamic Simulation Modeling Applications in Health Care Delivery Research Emerging Good Practices introduced the fundamentals of dynamic simulation modeling and identified the types of health care delivery problems for which dynamic simulation modeling can be used more effectively than other modeling methods. The hierarchical relationship between the health care delivery system, providers, patients, and other stakeholders exhibits a level of complexity that ought to be captured using dynamic simulation modeling methods. As a tool to help researchers decide whether dynamic simulation modeling is an appropriate method for modeling the effects of an intervention on a health care system, we presented the System, Interactions, Multilevel, Understanding, Loops, Agents, Time, Emergence (SIMULATE) checklist consisting of eight elements. This report builds on the previous work, systematically comparing each of the three most commonly used dynamic simulation modeling methods-system dynamics, discrete-event simulation, and agent-based modeling. We review criteria for selecting the most suitable method depending on 1) the purpose-type of problem and research questions being investigated, 2) the object-scope of the model, and 3) the method to model the object to achieve the purpose. Finally, we provide guidance for emerging good practices for dynamic simulation modeling in the health sector, covering all aspects, from the engagement of decision makers in the model design through model maintenance and upkeep. We conclude by providing some recommendations about the application of these methods to add value to informed decision making, with an emphasis on stakeholder engagement, starting with the problem definition. Finally, we identify areas in which further methodological development will likely occur given the growing "volume, velocity and variety" and availability of "big data" to provide empirical evidence and techniques such as machine learning for parameter estimation in dynamic simulation models. Upon reviewing this report in addition to using the SIMULATE checklist, the readers should be able to identify whether dynamic simulation modeling methods are appropriate to address the problem at hand and to recognize the differences of these methods from those of other, more traditional modeling approaches such as Markov models and decision trees. This report provides an overview of these modeling methods and examples of health care system problems in which such methods have been useful. The primary aim of the report was to aid decisions as to whether these simulation methods are appropriate to address specific health systems problems. The report directs readers to other resources for further education on these individual modeling methods for system interventions in the emerging field of health care delivery science and implementation.

Potential application of machine learning in health outcomes research and some statistical cautions

Traditional analytic methods are often ill-suited to the evolving world of health care big data characterized by massive volume, complexity, and velocity. In particular, methods are needed that can estimate models efficiently using very large datasets containing healthcare utilization data, clinical data, data from personal devices, and many other sources. Although very large, such datasets can also be quite sparse (e.g., device data may only be available for a small subset of individuals), which creates problems for traditional regression models. Many machine learning methods address such limitations effectively but are still subject to the usual sources of bias that commonly arise in observational studies. Researchers using machine learning methods such as lasso or ridge regression should assess these models using conventional specification tests.

Risk of stroke after catheter ablation versus cardioversion for atrial fibrillation: A propensity-matched study of 24,244 patients

BACKGROUND

Stroke is the major cause of morbidity and mortality related to atrial fibrillation (AF). Catheter ablation for AF is effective in reducing AF burden, but its impact on long-term stroke risk is unknown.

OBJECTIVE

We sought to evaluate the periprocedural and long-term stroke risk after catheter ablation or cardioversion for AF.

METHODS

This retrospective, propensity-matched study using a national administrative claims database identified patients with AF who underwent catheter ablation and a comparison group (matched on age, sex, year of treatment, CHA2DS2-Vasc score, and Charlson index) who underwent cardioversion between 2005 and 2012. The primary end points were (1) time to first ischemic or hemorrhagic stroke or transient ischemic attack (TIA) and (2) time to first ischemic or hemorrhagic stroke excluding TIA. We compared periprocedural incident stroke (within 30 days of ablation or cardioversion) as well as total strokes between the 2 groups.

RESULTS

A total of 24,244 patients (12,122 patients undergoing ablation and 12,122 patients undergoing cardioversion) were included in the analysis. Incident periprocedural stroke or TIA occurred in 0.5% of the ablation group and 0.3% of the cardioversion group (P = .04). There was a significant initial risk of stroke/TIA with ablation within the first 30 days (rate ratio 1.53; P = .05). After 30 days, this risk was significantly lower in the ablation group (rate ratio 0.78; P = .03).

CONCLUSION

In patients with AF, there is a small periprocedural stroke risk with ablation in comparison to cardioversion. However, over longer-term follow-up, ablation is associated with a slightly lower rate of stroke.

Use of the Optum Labs Data Warehouse to assess test ordering patterns for diagnosis of Helicobacter pylori infection in the United States

We surveyed national Helicobacter pylori diagnostic testing practices and diagnoses using commercial and Medicare medical claims data from Optum Labs (Cambridge, MA). Serologic testing for antibodies to H. pylori remains the most commonly ordered diagnostic test despite recent expert recommendations. Changes in reimbursement for serologic testing will likely drive future provider ordering practices.