A Case Study in

Real-world Evidence


Precision Analytics works with clients to plan and execute studies that reflect real-world interventions and populations. We draw from modern pharmacoepidemiology to identify approaches that fully leverage available data, while mitigating limitations to every extent possible.

Our clients rely on our work to support regulatory decision making, understand economic impacts, and demonstrate unmet needs.

Why real-world evidence?

While randomized controlled trials remain the “gold standard” for evaluating drug safety and efficacy, certain hypotheses are not possible, practical, or ethical to address experimentally. Trials also rely on strict selection criteria and protocols for internal validity, at the cost of reflecting everyday clinical practice.

  • Real-world evidence seeks to bridge the so-called “efficacy-effectiveness gap” that often arises between effects observed in clinical trials versus the real world.
  • Real-world data form the basis for real-world evidence and include a variety of non-experimental sources, typically collected during routine healthcare delivery.

Non-traditional study designs and data sources also offer new opportunities for monitoring drug safety after approval , and for understanding the full scope of patient outcomes. For instance, consumer wearables and health apps can provide a richer, more holistic understanding of patient experiences than conventional clinical assessments.

Examples of our work
1. Assessing the impact of screening criteria for gestational diabetes

The challenge

  • Gestational diabetes mellitus is a metabolic condition that can arise during pregnancy and increase the risk of dangerous complications.
  • We worked with investigators at the Montreal Lady Davis Institute for Medical Research (LDI) who were part of a collaboration to better understand the impact of gestational diabetes screening criteria on maternal and neonatal outcomes.

How we responded

  • We worked with investigators to develop and carry out an analysis plan that would take advantage of the existing differences between screening criteria used by local university-affiliated hospitals.
  • This work is helping clinicians to understand how differences in routine screening can affect key outcomes in childbirth such as abnormally high birth weight.
2. Outcomes after clinical support for intracerebral hemorrhage

The challenge

  • Our client, a researcher at a major neurological institute, wanted to measure mortality and functioning in a real-world cohort of stroke patients following medical and surgical interventions.
  • They were concerned about underlying differences between intervention groups, and wanted to ensure that their study was sufficiently large to provide clinically meaningful insights.

How we responded

  • We assessed study feasibility by conducting sample size calculations and proposed a statistical analysis plan incorporating propensity score matching to address potential confounding.
  • The final analysis examined a retrospective case series from tertiary neurocritical care units, and was led by one of our senior data scientists.
3. Cumulative anticholinergic exposure and risk of falls and fractures

The challenge

  • Older adults experience significant morbidity related to falls and fractures, for which anticholinergic use is a recognized risk factor. Existing trial data were insufficient to understand the magnitude of this association in the population of interest.
  • Clinical experts suspected a complex relationship between fall risk and anticholinergic exposure, which could vary and accumulate over time.

How we responded

  • We analyzed claims data for a U.S. commercially and Medicare-insured population relevant to the studied indication.
  • This retrospective cohort study demanded sophisticated handling of treatment patterns and the application of marginal structural models to address time-varying treatment exposure.
Call to action

Let us help with assessing and applying a variety of real-world data sources such as:

  • Pragmatic or open-label trials
  • Patient registries
  • Insurance claims
  • Electronic medical or health records (EMR/EHR)
  • Patient and physician surveys
  • Adverse event reporting databases (e.g., FAERS )
  • Wearable devices and biometrics

Our clients rely on us to identify the right data sources and approaches for their research questions, looking beyond the scope of clinical trials.


Let’s get work together

Contact us to reduce your data analysis workload and optimize your workflow.

Contact Us