RESEARCH DISCIPLINE
Statistical Analysis for Pharmacy & Clinical Pharmacology Research
Consulting for Medication Use & Outcomes Studies
Pharmacy research statistical analysis for dispensing and claims data, adherence studies, new-user cohorts and pharmacy practice trials, with exposure defined from the records, confounding by indication addressed and results prepared to the reporting standard your journal expects.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
Statistical Analysis for Pharmacy & Clinical Pharmacology Research
Pharmacy research statistical analysis is the statistical design, analysis and reporting of studies of how medicines are dispensed, taken and tolerated in patients, using pharmacy dispensing files, claims, health records, safety reports and pharmacy practice trials. Exposure has to be inferred from what was supplied, patients receive a medicine because of their condition, and dose and kidney function change over time, so a workflow that compares users with non-users at one baseline gives biased and over-confident estimates. OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, provides pharmacy data analysis through in-house PhD-level statisticians, who work out the exposure definition, the comparator and the start of follow-up with you before any outcome model is run.
WHEN TO ENGAGE
Statistical Support at Different Stages of Pharmacy & Pharmacology Research
Pharmacy researchers contact us from writing the exposure definition of a protocol to answering a reviewer's question about confounding in a finished cohort.
Defining Exposure and Comparators You are planning a drug utilization or safety study and need the exposure definition, the comparator, the grace period and the start of follow-up written down before the dispensing file is queried.
Analyzing Dispensing or Claims Data Your extract is ready, and the adherence measure, the outcome model and the treatment of changing dose and hospital stays need to be chosen, run and checked.
Evaluating a Pharmacy Service A medication review, counseling or refill program has been delivered in one or several pharmacies, and the outcomes of its patients need an analysis that accounts for the pharmacy they used.
Preparing a Pharmacoepidemiology Paper Results are ready to be written up and need adjusted estimates with intervals, cohort flow and a methods section prepared to the reporting guidance your target journal expects.
RESEARCH CONTEXT
Research Designs and Data We Work With in Pharmacy & Pharmacology
Pharmacy data come from dispensing systems, claims and clinical monitoring, and each source records exposure differently from what a trial would.
01
Dispensing and Prescription Databases
Fill records from pharmacies, claims or health plans, where exposure is inferred from days of supply and follow-up is set by enrollment and gaps.
02
Adherence and Persistence Studies
Proportion of days covered, medication possession and time to discontinuation, measured from refills or electronic monitors, with gaps and stockpiling to define.
03
Drug Utilization and Safety Cohorts
New-user cohorts comparing medicines in patients treated for clinical reasons, with doses that change and covariates such as kidney function that vary over time.
04
Pharmacy Practice Interventions
Medication review, counseling or service programs evaluated in one or several pharmacies, with patients nested in pharmacies and outcomes counted per patient.
05
Pharmacovigilance Reports
Spontaneous adverse event reports and safety signals, where there is no count of exposed patients and reporting itself varies with attention to a drug.
06
Drug Levels and Dose-Response in Patients
Concentrations from therapeutic drug monitoring, doses and clinical response measured repeatedly in the same patients, with kidney or liver function that changes between visits.
METHODOLOGICAL CONSIDERATIONS
Statistical Issues We Consider in Pharmacy & Pharmacology Research
Exposure, comparator and outcome measures are examined before results are trusted, since each can shift an estimate further than the choice of model.
Exposure Inferred From Fills A fill is not a dose taken. Days of supply, grace periods and carryover of leftover supply are set in advance and varied in sensitivity analyses, since adherence estimates shift with each choice.
New Users and Active Comparators Prevalent users have already survived and tolerated the early months of treatment, so follow-up starts at the first fill after a washout period, with patients starting another active treatment as the comparator.
Confounding by Indication Medicines are given because of the condition they treat, so unadjusted differences between treatments can largely reflect who was treated. Covariates come from a causal argument, with an active comparator preferred.
Time-Varying Dose and Kidney Function Kidney function guides dosing and may be changed by it, so time-varying adjustment in a standard model can be biased; marginal structural models are considered, and dose effects are not assumed to be linear.
Adherence Is Not a Normal Outcome Proportion of days covered is bounded between zero and one, with a spike at one. It is modeled as a proportion or time to discontinuation, and a cutoff such as 0.80 is shown beside the continuous result.
Signals Versus Risk Spontaneous reports give a signal, not an incidence, and disproportionality is not read as risk. Where exposed patients can be counted, rates with intervals are reported.
STATISTICAL METHODS
Methods Selected for the Research Question and Study Design
The method families below follow the outcome and the exposure structure of your study, and each line says why a family suits medication data.
Clinical, Survival & Epidemiological
Time to discontinuation or an adverse event runs from each patient's first fill, and propensity methods balance the measured reasons for choosing one medicine over another.
Regression & Predictive Modeling
Adherence above a threshold, adverse event counts and dispensing frequency each receive a model for their own outcome type, with person-time as the offset where events are counted.
Longitudinal & Multilevel Analysis
Drug levels and kidney function measured at repeated visits, and patients nested in pharmacies or prescribers, call for models that represent the clustering.
Multivariate, Scale & Reliability
Self-reported adherence and beliefs-about-medicines scales are checked for reliability and scoring before they are set against refill-based measures.
Comparative & Hypothesis Testing
The same patients meeting an adherence target before and after a pharmacy service are compared with McNemar's test, and paired change in a continuous score with a paired t-test.
Bayesian & Computational Methods
Sparse adverse event reports are screened with Bayesian shrinkage, and simulation tests a cohort estimate against plausible misclassification of fill-based exposure.
DELIVERABLES
Statistical Outputs Prepared for Your Research
The outputs follow the structure of a medication study, from the review of the exposure algorithm to the programs that rebuild the cohort and every reported value.
Data & Study Review The dispensing or claims dictionary, days-of-supply rules, enrollment windows, exposure and comparator definitions and time-varying covariates are reviewed before the cohort is built.
Statistical Analysis The agreed adherence, survival, count or mixed models, with balance checks for treatment comparisons and sensitivity analyses on the exposure definition and the grace period.
Effect Estimates & Uncertainty Hazard, rate or odds ratios with absolute risk differences and confidence intervals, and adherence summarized as a continuous measure and at the threshold you require.
Tables & Figures Cohort flow with exclusions, baseline tables before and after weighting or matching, balance plots, adherence trajectories and time-to-discontinuation curves.
Statistical Interpretation Findings are worded as associations between a medicine and an outcome, confounding by indication and exposure misclassification are discussed, and relevance is given in clinical terms.
Analysis Files Programs in SAS, R, Stata or SPSS, with the exposure algorithm documented step by step so that the cohort can be rebuilt from the dispensing file.
REPORTING
Reporting Pharmacy & Pharmacology Research
A medication study follows the guideline for its design, with extensions for dispensing and claims records and for the way adherence was measured.
Cohort and case-control studies of medicines follow STROBE, with RECORD and its pharmacoepidemiology extension RECORD-PE where the data are dispensing, claims or health records. The protocol of such a study can be set out in the HARPER template, and studies of adherence describe how it was measured and reported following the EMERGE guideline. Randomized trials of a medicine or a pharmacy service report to CONSORT 2025. Exposure definitions, grace periods and comparator logic are tabulated so reviewers can reproduce the cohort, adjusted ratios are accompanied by absolute risks, and the instructions of your target journal are checked at the end.
ILLUSTRATIVE EXAMPLE
How the Analysis May Look in Practice
An illustration of how an adherence question in dispensing records moves from an exposure definition to a result that withstands review.
ILLUSTRATIVE EXAMPLE · COMMUNITY PHARMACY ADHERENCE RESEARCH · UNITED STATES
A community pharmacy research network wanted to know whether enrollment in a refill-synchronization program was associated with better adherence to statins, using dispensing records from several pharmacies.
Patients enrolled months after their first fill, so comparing enrolled patients with others over the whole year would credit the program with time before enrollment. Before outcomes were examined, we set a landmark six months after the first fill, defined groups by enrollment status at the landmark, adjusted for adherence and covariates measured before it, and measured adherence afterward, with hospital days removed from the supply calculation. Adherence was modeled as a proportion with generalized estimating equations clustered by pharmacy, and the proportion at or above 0.80 was reported beside it. The grace period was varied in a sensitivity analysis.
Outcome: The group received adjusted differences in proportion of days covered and in the proportion at or above 0.80, each with its 95% confidence interval, the grace-period sensitivity results, RECORD-PE flow counts, a methods paragraph and the scripts.
This example does not describe a real client or project. We never share details of actual client work, in line with the confidentiality we promise every researcher. This illustrative example is included to show the type, scope, and standard of work involved in this kind of analysis.
HOW WE ASSESS YOUR PROJECT
We Review Your Actual Research
Your assessment is based on a review of your actual research materials and analytical requirements, not a generic project category.
You Submit Your Project Details
Share whatever is available at the current stage, such as the research aim, protocol or proposal, dataset, variable list, manuscript or statistical requirements.
We Review Your Research and Define the Scope
Your statistically trained Research Coordinator reviews the materials to understand the study design, research questions, data structure, analytical requirements and expected deliverables. If clarification is needed, additional documents or information can be requested before the assessment is finalized.
Your Analysis Is Completed and Delivered
Once the scope and fixed project fee are agreed, the in-house statistician completes the work and the Research Coordinator remains the communication point through delivery, with the analysis, deliverables and timeline as agreed.
SOFTWARE
Statistical Software for Pharmacy & Pharmacology Research
Medication analyses are run in the software that suits claims and dispensing files, with syntax or scripts provided where applicable.
Statistical Support Within Your Research OnlineSPSS provides the agreed methodological, analytical, and statistical reporting support. The researcher or research team retains the scientific question, substantive interpretation, conclusions, and overall responsibility for the research.
FREQUENTLY ASKED QUESTIONS
Questions About Statistical Analysis for Pharmacy & Pharmacology
Answers to common questions from pharmacy researchers, on adherence measures, new-user cohorts, confounding, review, confidentiality and the assessment.
How is medication adherence analyzed statistically?
In medication adherence analysis, adherence from refill records is usually a proportion of days covered or a time to discontinuation, so it is modeled as a bounded proportion, a binary threshold reported beside the continuous result, or a time-to-event outcome, not as a normally distributed score. The days-of-supply rules and grace period are fixed first and varied in sensitivity analyses. Our Statistical Methods pages describe the families involved.
Can you analyze dispensing or claims data as a new-user cohort?
Yes. The cohort is built from the dispensing and enrollment files: a washout window confirms that each patient is new to the drug, continuous enrollment is checked before and after the first fill, and every exclusion is counted in the cohort flow, so a reviewer can see who entered and why.
How do you handle confounding when comparing two medicines?
Confounding by indication is addressed by design first, through new users and an active comparator, and by adjustment second. Covariates are chosen from a causal argument about what influenced both prescribing and the outcome, balance is checked after weighting or matching, and residual confounding is stated as a limitation. Results are worded as associations, not effects.
Can you review a pharmacoepidemiology analysis before submission?
We can review a completed analysis before submission or in response to reviewer comments. The review covers the exposure definition, the start of follow-up, immortal time, the covariate set, the handling of time-varying dose and the reporting against STROBE and RECORD-PE, and states what stands and what should change.
How is confidential medication and patient data handled?
Study datasets, protocols and analysis files are handled confidentially, and project information is not published. An NDA is available where required and can be arranged before sensitive materials are shared. Read more about confidentiality and data security.
What happens when I request an assessment, and how long does it take?
You share your research question, design and relevant materials, and a statistically trained Research Coordinator reviews the request and responds the same day. The analysis, deliverables, timeline and one fixed project fee are then confirmed before any work begins, so the timeline follows from the agreed scope.
Still have a question? Find more answers in our full FAQ.
NEXT STEP
Discuss Your
Pharmacy & Pharmacology Research With Us
Share your research question, study design and available data, even if only a protocol or a data dictionary exists. We review how exposure is recorded and confirm the approach before any work begins.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014