Linking electronic health data in pharmacoepidemiology: Appropriateness and feasibility.

PURPOSE: To provide guidance on data linkage appropriateness and feasibility to plan purposeful and sustainable new linkages that advance pharmacoepidemiology and healthcare research. Planning a new data linkage requires careful evaluation to weigh the resources required with the potential overall benefits. METHODS: In response to an International Society for Pharmacoepidemiology (ISPE) call for manuscripts, a working group comprised of members from academic, industry, and government determined priority content areas; appropriateness and feasibility of data linkage was selected. Within this topic, scientific and operational considerations were determined, reviewed, and formulated into key areas, and translated into 12 consensus recommendations. RESULTS: Guidance for feasibility assessment was categorized into five key areas: (1) research objectives and justification; (2) data quality and completeness; (3) the linkage process; (4) data ownership and governance; and (5) overall value added by linkage. Within these key areas, recommendations to consider prior to initiation were developed to evaluate suitability of the linkage to meet research objectives, assess source data completeness and population coverage, and ensure well-defined data governance standards and protections. When creating novel linked datasets, researchers must assess the feasibility of both scientific (data quality and linkage methods) and operational (access, data use and transfer, governance, and cost) aspects. CONCLUSIONS: The data linkage feasibility assessment considerations outlined can be used as a guide when designing sustainable linked data resources to generate actionable evidence in healthcare research. These recommendations were constructed for wide applicability and can be adapted depending on the geographic, structural, and data components of the linkage.

Validity of diagnostic codes to identify hospitalizations for infections among patients treated with oral anti-diabetic drugs.

PURPOSE: Identification of hospitalizations for infection is important for post-marketing surveillance of drugs, but the validity of using diagnosis codes to identify these events is unknown. Differentiating between hospitalization for and with infection is important, as the latter is common and less likely to arise from pre-admission exposure to drugs. We determined positive predictive values (PPVs) of diagnostic coding-based algorithms to identify hospitalization for infection among patients prescribed oral anti-diabetic drugs (OADs). METHODS: We identified patients initiating OADs within 2 United States claims databases (Medicare, HealthCore Integrated Research DatabaseSM [HIRDSM ]) and 2 United Kingdom electronic medical record databases (Clinical Practice Research Datalink [CPRD], The Health Improvement Network [THIN]) from 2009 to 2014. To identify potential hospitalizations for infection, we selected patients with a hospital diagnosis of infection and, within 7 days prior to hospitalization, either an outpatient/emergency department visit with an infection diagnosis or outpatient antimicrobial treatment. Hospital records were reviewed by infectious disease specialists to adjudicate hospital admissions for infection. PPVs for confirmed outcomes were determined for each database. RESULTS: Code-based algorithms to identify hospitalization for infection had PPVs exceeding 80% within Medicare (PPV, 83% [90/109]; 95% CI, 74-89%), HIRDSM (PPV, 89% [73/82]; 95% CI, 80-95%), and THIN (PPV, 86% [12/14]; 95% CI, 57-98%) but not within CPRD (PPV, 67% [14/21]; 95% CI, 43-85%). CONCLUSIONS: Algorithms identifying hospitalization for infection utilizing hospital diagnoses along with antecedent outpatient/emergency infection diagnoses or antimicrobial therapy had sufficiently high PPVs for confirmed events within Medicare, HIRDSM , and THIN to enable their use for pharmacoepidemiologic research.

Evaluation of methods to estimate missing days' supply within pharmacy data of the Clinical Practice Research Datalink (CPRD) and The Health Improvement Network (THIN).

PURPOSE: The extent to which days' supply data are missing in pharmacoepidemiologic databases and effective methods for estimation is unknown. We determined the percentage of missing days' supply on prescription and patient levels for oral anti-diabetic drugs (OADs) and evaluated three methods for estimating days' supply within the Clinical Practice Research Datalink (CPRD) and The Health Improvement Network (THIN). METHODS: We estimated the percentage of OAD prescriptions and patients with missing days' supply in each database from 2009 to 2013. Within a random sample of prescriptions with known days' supply, we measured the accuracy of three methods to estimate missing days' supply by imputing the following: (1) 28 days' supply, (2) mode number of tablets/day by drug strength and number of tablets/prescription, and (3) number of tablets/day via a machine learning algorithm. We determined incidence rates (IRs) of acute myocardial infarction (AMI) using each method to evaluate the impact on ascertainment of exposure time and outcomes. RESULTS: Days' supply was missing for 24 % of OAD prescriptions in CPRD and 33 % in THIN (affecting 48 and 57 % of patients, respectively). Methods 2 and 3 were very accurate in estimating days' supply for OADs prescribed at a consistent number of tablets/day. Method 3 was more accurate for OADs prescribed at varying number of tablets/day. IRs of AMI were similar across methods for most OADs. CONCLUSIONS: Missing days' supply is a substantial problem in both databases. Method 2 is easy and very accurate for most OADs and results in IRs comparable to those from method 3.

Use of demographic and pharmacy data to identify patients included within both the Clinical Practice Research Datalink (CPRD) and The Health Improvement Network (THIN).

PURPOSE: Pharmacoepidemiology researchers often utilize data from two UK electronic medical record databases, the Clinical Practice Research Datalink (CPRD) and The Health Improvement Network (THIN), and may choose to combine the two in an effort to increase sample size. To minimize duplication of data, previous studies examined the practice-level overlap between these databases. However, the proportion of overlapping patients remains unknown. We developed a method using demographic and pharmacy variables to identify patients included in both CPRD and THIN, and applied this method to measure the proportion of overlapping patients who initiated the oral anti-diabetic drug saxagliptin. METHODS: We conducted a cross-sectional study among patients initiating saxagliptin in CPRD and THIN between October 2009 and September 2012. Within both databases, we identified patients: (i) ≥18 years, (ii) newly prescribed saxagliptin, and (iii) with ≥180 days enrollment prior to saxagliptin initiation. Demographic data (birth year, sex, patient registration date, family number, and marital status) and prescriptions (including dates) for the first two oral anti-diabetic drugs prescribed within the study period were used to identify matching patients. RESULTS: Among 4202 CPRD and 3641 THIN patients initiating saxagliptin, 2574 overlapping patients (61% of CPRD saxagliptin initiators; 71% of THIN saxagliptin initiators) were identified. Among these patients, 2474 patients (96%) perfectly matched on all demographic and prescription data. CONCLUSIONS: Within each database, over 60% of patients initiating saxagliptin were included within both CPRD and THIN. Combined demographic and prescription data can be used to identify patients included in both CPRD and THIN.

Safety of saxagliptin: rationale for and design of a series of postmarketing observational studies.

PURPOSE: To describe the design and rationale of a series of postmarketing studies to examine the safety of saxagliptin, an oral dipeptidyl peptidase-4 inhibitor for the treatment of type 2 diabetes mellitus, in real-world settings. METHODS: We are conducting a series of retrospective cohort studies using two UK (General Practice Research Database, and The Health Improvement Network) and two US (Medicare, HealthCore Integrated Research Database(SM) ) data sources. The primary outcomes of interest will include (i) hospitalization with acute liver failure, (ii) hospitalization for acute kidney injury, (iii) hospitalization for severe hypersensitivity reactions, (iv) hospitalization for severe infections, (v) hospitalization with infections associated with T-lymphocyte dysfunction (i.e., herpes zoster, tuberculosis, or nontuberculous mycobacteria), and (vi) major cardiovascular events. Diagnosis codes for the outcomes of interest will be validated by medical record review within each data source. Projected use and estimated incidence rates of outcomes of interest suggest there will be at least 80% statistical power to detect a minimum hazard ratio of 1.5 for major cardiovascular events, 2.0 for acute kidney injury and severe infections, 2.4 for acute liver failure, and 4.0 for severe hypersensitivity reactions. RESULTS: Forthcoming. CONCLUSIONS: This postmarketing safety assessment will provide important information regarding the safety of saxagliptin and could potentially identify important dipeptidyl peptidase-4 inhibitor class effects. The methods described may be useful to others planning similar evaluations.

Postauthorization safety study of the DPP-4 inhibitor saxagliptin: a large-scale multinational family of cohort studies of five outcomes.

OBJECTIVE: To evaluate the risk of serious adverse events among patients with type 2 diabetes mellitus initiating saxagliptin compared with oral antidiabetic drugs (OADs) in classes other than dipeptidyl peptidase-4 (DPP-4) inhibitors. RESEARCH DESIGN AND METHODS: Cohort studies using 2009-2014 data from two UK medical record data sources (Clinical Practice Research Datalink, The Health Improvement Network) and two USA claims-based data sources (HealthCore Integrated Research Database, Medicare). All eligible adult patients newly prescribed saxagliptin (n=110 740) and random samples of up to 10 matched initiators of non-DPP-4 inhibitor OADs within each data source were selected (n=913 384). Outcomes were hospitalized major adverse cardiovascular events (MACE), acute kidney injury (AKI), acute liver failure (ALF), infections, and severe hypersensitivity events, evaluated using diagnostic coding algorithms and medical records. Cox regression was used to determine HRs with 95% CIs for each outcome. Meta-analyses across data sources were performed for each outcome as feasible. RESULTS: There were no increased incidence rates or risk of MACE, AKI, ALF, infection, or severe hypersensitivity reactions among saxagliptin initiators compared with other OAD initiators within any data source. Meta-analyses demonstrated a reduced risk of hospitalization/death from MACE (HR 0.91, 95% CI 0.85 to 0.97) and no increased risk of hospitalization for infection (HR 0.97, 95% CI 0.93 to 1.02) or AKI (HR 0.99, 95% CI 0.88 to 1.11) associated with saxagliptin initiation. ALF and hypersensitivity events were too rare to permit meta-analysis. CONCLUSIONS: Saxagliptin initiation was not associated with increased risk of MACE, infection, AKI, ALF, or severe hypersensitivity reactions in clinical practice settings. TRIAL REGISTRATION NUMBER: NCT01086280, NCT01086293, NCT01086319, NCT01086306, and NCT01377935; Results.

Determinants of saxagliptin use among patients with type 2 diabetes mellitus treated with oral anti-diabetic drugs.

BACKGROUND: The patterns and determinants of saxagliptin use among patients with type 2 diabetes mellitus (T2DM) are unknown in real-world settings. We compared the characteristics of T2DM patients who were new initiators of saxagliptin to those who were new initiators of non-dipeptidyl peptidase-4 (DPP-4) inhibitor oral anti-diabetic drugs (OADs) and identified factors associated with saxagliptin use. METHODS: We conducted a cross-sectional study within the Clinical Practice Research Datalink (CPRD), The Health Improvement Network (THIN), US Medicare, and the HealthCore Integrated Research Database (HIRD(SM)) across the first 36 months of saxagliptin availability (29 months for US Medicare). Patients were included if they were: 1) ≥18 years old, 2) newly prescribed saxagliptin or a non-DPP-4 inhibitor OAD, and 3) enrolled in their respective database for 180 days. For each saxagliptin initiator, we randomly selected up to ten non-DPP-4 inhibitor OAD initiators matched on age, sex, and geographic region. Conditional logistic regression was used to identify determinants of saxagliptin use. RESULTS: We identified 64,079 saxagliptin initiators (CPRD: 1,962; THIN: 2,084; US Medicare: 51,976; HIRD(SM): 8,057) and 610,660 non-DPP-4 inhibitor OAD initiators (CPRD: 19,484; THIN: 19,936; US Medicare: 493,432; HIRD(SM): 77,808). Across all four data sources, prior OAD use, hypertension, and hyperlipidemia were associated with saxagliptin use. Saxagliptin initiation was also associated with hemoglobin A1c results >8% within the UK data sources, and a greater number of hemoglobin A1c measurements in the US data sources. CONCLUSIONS: In these UK and US data sources, initiation of saxagliptin was associated with prior poor glycemic control, prior OAD use, and diagnoses of hypertension and hyperlipidemia. TRIAL REGISTRATION: ClinicalTrials.gov identifiers NCT01086280 , NCT01086293 , NCT01086319 , NCT01086306 , and NCT01377935.

An evaluation of the THIN database in the OMOP Common Data Model for active drug safety surveillance.

BACKGROUND: There has been increased interest in using multiple observational databases to understand the safety profile of medical products during the postmarketing period. However, it is challenging to perform analyses across these heterogeneous data sources. The Observational Medical Outcome Partnership (OMOP) provides a Common Data Model (CDM) for organizing and standardizing databases. OMOP's work with the CDM has primarily focused on US databases. As a participant in the OMOP Extended Consortium, we implemented the OMOP CDM on the UK Electronic Healthcare Record database-The Health Improvement Network (THIN). OBJECTIVE: The aim of the study was to evaluate the implementation of the THIN database in the OMOP CDM and explore its use for active drug safety surveillance. METHODS: Following the OMOP CDM specification, the raw THIN database was mapped into a CDM THIN database. Ten Drugs of Interest (DOI) and nine Health Outcomes of Interest (HOI), defined and focused by the OMOP, were created using the CDM THIN database. Quantitative comparison of raw THIN to CDM THIN was performed by execution and analysis of OMOP standardized reports and additional analyses. The practical value of CDM THIN for drug safety and pharmacoepidemiological research was assessed by implementing three analysis methods: Proportional Reporting Ratio (PRR), Univariate Self-Case Control Series (USCCS) and High-Dimensional Propensity Score (HDPS). A published study using raw THIN data was selected to examine the external validity of CDM THIN. RESULTS: Overall demographic characteristics were the same in both databases. Mapping medical and drug codes into the OMOP terminology dictionary was incomplete: 25 % medical codes and 55 % drug codes in raw THIN were not listed in the OMOP terminology dictionary, representing 6 % condition occurrence counts, 4 % procedure occurrence counts and 7 % drug exposure counts in raw THIN. Seven DOIs had <0.3 % and three DOIs had 1 % of unmapped drug exposure counts; each HOI had at least one definition with no or minimal (≤0.2 %) issues with unmapped condition occurrence counts, except for the upper gastrointestinal (UGI) ulcer hospitalization cohort. The application of PRR, USCCS and HDPS found, respectively, a sensitivity of 67, 78 and 50 %, and a specificity of 68, 59 and 76 %, suggesting that safety issues defined as known by the OMOP could be identified in CDM THIN, with imperfect performance. Similar PRR scores were produced using both CDM THIN and raw THIN, while the execution time was twice as fast on CDM THIN. There was close replication of demographic distribution, death rate and prescription pattern and trend in the published study population and the cohort of CDM THIN. CONCLUSIONS: This research demonstrated that information loss due to incomplete mapping of medical and drug codes as well as data structure in the current CDM THIN limits its use for all possible epidemiological evaluation studies. Current HOIs and DOIs predefined by the OMOP were constructed with minimal loss of information and can be used for active surveillance methodological research. The OMOP CDM THIN can be a valuable tool for multiple aspects of pharmacoepidemiological research when the unique features of UK Electronic Health Records are incorporated in the OMOP library.