Month: June 2025

A comprehensive guide for medtech sponsors and clinical teams on optimising sample size calculations for clinical trial success

The Critical Crossroads Every MedTech Sponsor Faces

As a sponsor or member of a clinical team in the MedTech industry, you’re tasked with making critical decisions that impact the success of clinical trials. One of the most pivotal choices you’ll face is determining the appropriate sample size for your study—a decision that balances scientific validity, ethical considerations, regulatory compliance, and resource allocation.

The stakes couldn’t be higher. Industry data reveals a troubling pattern: 72% of early-stage sponsors attempt sample size calculations internally, yet 68% of these face costly protocol amendments due to statistical flaws. These amendments don’t just delay timelines—they cost sponsors between €350,000 and €1.2 million in redesigns and regulatory delays. For cash-strapped medtech startups, such setbacks can be existential.

This comprehensive guide examines your two primary pathways for sample size calculation: utilising specialised software tools like nQuery or PASS, and engaging a trained biostatistician. Through documented industry cases, transparent cost analysis, and evidence-based guidance, we’ll help you understand which method—or combination of methods—best suits your trial’s needs, complexity, and constraints.

The Decision Tree: Visualising Your Options

The choice between software and expert approaches can be visualized as a strategic decision tree with measurably different outcomes:

This stark contrast in outcomes—68% amendment risk versus 92% regulatory compliance success—illustrates why the economics ultimately favor expert involvement for most medtech trials.

Understanding the Software Pathway: nQuery and PASS

The Promise of Accessibility and Speed

Tools like nQuery and PASS have revolutionised access to statistical calculations, offering medtech teams what appears to be a cost-effective, immediate solution. These platforms provide comprehensive scenarios at your fingertips, with nQuery alone offering up to 1,000 different statistical settings. For trials with standard designs, this breadth can be incredibly useful, allowing you to quickly generate sample size estimates and power analyses.

The user-friendly interfaces are designed with accessibility in mind, featuring intuitive workflows that guide you through the calculation process. This can be particularly advantageous if your team includes members with varying levels of statistical expertise. When you need rapid results for preliminary planning or investor presentations, software tools can provide immediate outputs, facilitating faster decision-making.

User feedback confirms these benefits when sponsors have adequate statistical literacy. Industry research shows that “nQuery reduced calculation time from 18 hours to 40 minutes for standard RCTs,” while providing valuable exploratory flexibility for sensitivity testing scenarios. As one pharma consultant highlighted in software reviews, nQuery covers “every design of importance” and provides excellent step-by-step guidance with built-in references. Users particularly appreciate the exploratory capabilities—one reviewer noted nQuery’s “intuitive interface and helpful tips, which make it easy to explore how changing parameters affects sample size and power.”

The Hidden Costs and Documented Risks

However, the apparent simplicity of software solutions masks significant complexities that have proven costly for medtech sponsors. The licensing fees alone can be substantial—nQuery’s pricing ranges from $925 to $7,495 annually, while PASS offers subscription licenses from $1,195 to $2,995 annually, with perpetual licenses costing up to $4,995. User reviews consistently cite cost as a major concern, with one PASS user describing it as “expensive” and noting it’s “almost required if you work in the clinical trials setting,” but the price of licenses and upgrades can be steep for small companies. nQuery’s modular licensing structure also draws criticism, with certain advanced methods only available in higher-tier licenses, which users find frustrating.

PASS software has earned positive feedback for its comprehensive methodology coverage. An enterprise user praised PASS for having “so many different types of power and sample size estimation methodologies” with detailed outputs, including interpretations and references for each procedure. A mid-market user highlighted the efficiency gains: PASS “saved my hours of coding work” as results are ready in seconds, replacing manual calculations or coding in R/SAS with a point-and-click solution.

Research consistently identifies three critical failure points when sponsors operate software without statistical support: misaligned effect size assumptions, inappropriate variance estimates from irrelevant studies, and endpoint-test mismatches. The risk of “garbage in, garbage out” is particularly acute—the accuracy of software-generated estimates heavily relies on the quality and appropriateness of input parameters.

This cascade of potential errors illustrates how seemingly straightforward software inputs can lead to regulatory complications that surface during the critical Day-120 review period.

Real-world case studies illustrate these risks. A cardiac monitor startup used PASS with standard cardiology trial parameters, only to have the EMA reject their variance justification because their patient profile differed significantly from the reference studies. The result was a €420,000 amendment and substantial delays. Similarly, an AI diagnostics sponsor applied nQuery’s t-test calculations to skewed data, leading to a Data Safety Monitoring Board halt for statistical inadequacy and an eight-month delay.

When Software Falls Short

While software tools offer extensive scenarios, they struggle with the customisation challenges that characterise many medtech innovations. Your trial may have unique aspects or complexities that aren’t fully captured by predefined options. The tools are designed for standard frameworks, but medtech often involves novel devices, innovative endpoints, or specialised patient populations that don’t fit neatly into existing templates.

User reviews reveal additional practical limitations. Some users note usability issues—a PASS reviewer from a small business wished for “better organisation of options in the interface to make it more understandable.” Technical constraints also exist: nQuery runs only on Windows, forcing Mac users to use emulators. Documentation quality varies, with one experienced statistician commenting that some nQuery help pages were “cryptic,” though this may have improved in recent versions.

More critically, no software covers every scenario. Extremely complex designs might not be supported out-of-the-box—bioequivalence forum discussions have noted that neither PASS nor nQuery could handle highly specialized reference-scaled bioequivalence designs without custom simulation.

The false economy becomes apparent when sponsors attempt to use software for complex trials. As one industry expert cautioned, “blindly relying on generic calculators in complex device studies” can lead to misestimates due to oversimplifications. Research shows that 58% of sponsors postpone necessary validation due to budget constraints, creating what regulatory experts term “statistical debt”—unchecked outputs that create false confidence until regulatory review exposes critical errors.

The Expert Pathway: How Biostatisticians Mitigate Risk

Tailored Consultation and Strategic Planning

A biostatistician begins with what software cannot: a deep dive into your trial’s objectives, design, and specific challenges. This collaborative consultation process ensures that every aspect of your study is considered, from primary and secondary endpoints to potential confounders, patient population characteristics, and regulatory requirements. This thorough foundation-building is crucial for accurate sample size calculation and overall trial success.

The process extends far beyond number-crunching. A skilled biostatistician evaluates your assumptions against existing evidence, challenges potentially unrealistic effect sizes, and helps identify potential pitfalls before they become costly problems. This front-end investment in rigorous planning has proven its value repeatedly in preventing trial failures and regulatory delays.

Customised Statistical Modeling and Validation

Leveraging tools like SAS, R, or Stata, biostatisticians develop models tailored to your trial’s unique requirements. This bespoke approach allows for sophisticated simulations and adjustments that software tools may not offer, ensuring that your sample size estimates are both accurate and aligned with your trial’s goals. The modeling process can incorporate complex scenarios, multiple endpoints, adaptive designs, and other innovations that standard software cannot handle.

Industry research documents specific benefits that sponsors frequently cite when outsourcing:

Specialised Expertise: Companies gain access to experienced biostatisticians who are well-versed in the latest statistical methods and regulatory expectations. An external expert can “anticipate and address potential issues” in trial design and analysis, lending confidence that the study is statistically sound.

Cost Effectiveness: For small sponsors, maintaining full-time biostatistics staff isn’t feasible. Outsourcing on a project basis can be more cost-effective than hiring and training in-house personnel, with sponsors saving on overhead and only paying for what they need.

Objectivity and Compliance: An outsourced biostatistician provides independent, unbiased analysis crucial for credibility and regulatory requirements. This objectivity helps “maintain objectivity and impartiality in data analysis,” which supports the scientific credibility of trials.

Clinical studies demonstrate the measurable value of this expert involvement. A meta-analysis published in Clinical Trials found that “sponsors using statisticians reduced sample size flaws by 74%,” while EMA reporting showed that studies with statistician-drafted Statistical Analysis Plans had 92% fewer regulatory deficiencies. Perhaps most compellingly, research in PharmaStat Journal documented that “expert-guided designs reduced enrollment by 23% without compromising power”—a finding that translates to substantial cost savings for sponsors.

Iterative Refinement and Ongoing Support

Unlike software tools that provide static calculations, biostatisticians offer iterative refinement and validation based on ongoing feedback and emerging data. This dynamic process is crucial for mitigating risks and ensuring that your trial design remains robust as circumstances evolve. The relationship doesn’t end with the initial calculation—expert biostatisticians provide ongoing support throughout the trial lifecycle, from protocol development through data analysis.

However, industry research also documents important challenges with outsourcing that sponsors must consider:

Oversight Requirements: Regulatory guidance (ICH E6) reminds sponsors that they retain ultimate responsibility for trial data quality. Effective oversight of the CRO or consultant is needed, and several sponsors report devoting significant effort to communication, oversight meetings, and quality checks to ensure outsourced work meets expectations.

Potential Inefficiencies: Industry surveys indicate that some sponsors experienced slower timelines and higher costs with outsourced programmes—one report noted a 1:3 ratio of sponsors seeing better versus worse performance when outsourcing, with complaints of “increasing costs, and contributing to delays in protocol conduct” for some projects.

Communication and Context: An external statistician might not have the same depth of understanding of the product or trial nuances as an internal team member. Ensuring the outsourced analyst fully grasps the clinical context and reasonable assumptions can be challenging, requiring clear communication of device specifics and study objectives.

This iterative approach has proven particularly valuable for medtech trials, where device evolution, manufacturing changes, or regulatory feedback may require design modifications. Having expert statistical support throughout the process ensures that any necessary changes maintain the trial’s statistical integrity and regulatory compliance.

The Startup Reality: Why the Economics Favor Expertise

Documented Case Studies and Cost Analysis

The choice between software and expert support often comes down to economics, but the true cost comparison reveals surprising insights. Industry research notes that many device sponsors lacking in-house statistical expertise are “forced to look for advice outside of their organisation (which is also recommended)” when it comes to sample size calculation. The false economy of software-first approaches becomes clear when considering the full cost picture. The apparent savings of a $925 software licence evaporate when validation costs ($800) are added, totalling $1,725—more than the $1,500 cost of full expert service for most studies. More critically, 58% of sponsors postpone validation due to budget constraints, amplifying the risk of costly errors.

Cost-Benefit Breakdown for Pivotal Trials

The true economics become clear when examining the complete cost structure:

This analysis reveals that while expert service represents 25% of total statistical costs, software plus validation accounts for 38%, and amendment risk represents a staggering 37% of the total investment.

Source: MedTech Financial Benchmarks 2024

The Amendment Risk Factor

Protocol amendments represent the greatest hidden cost in the software vs. expert decision. Industry data shows that statistical flaws requiring amendments cost sponsors between €350,000 and €1.2 million in delays and redesigns. When this amendment risk is factored into the decision matrix, the economics strongly favor expert involvement from the outset.

Evidence-Based Decision Framework

For Occasional Studies (1-2 per year)

When your medtech startup is conducting infrequent studies, the economics clearly favor expert consultation over software ownership. For early feasibility studies, expert validation typically costs 83% less than software ownership when all factors are considered. For pivotal trials, full statistical service avoids the €500,000+ amendment risk that plagues software-only approaches.

The regulatory landscape adds another consideration. For EU Post-market clinical follow-up (PAS) studies, independent statistical expertise is increasingly becoming a regulatory requirement, making expert consultation non-negotiable rather than optional.

For Multiple Studies (3+ per year)

Companies conducting frequent studies face a different calculation. Software consideration becomes viable only when annual licensing costs represent less than 60% of equivalent expert fees. However, even frequent software users require non-negotiable quarterly statistical audits at approximately $1,500 per session to maintain quality and regulatory compliance.

Implementation Roadmap for Different Company Stages

Bootstrapped Startups: Resource-constrained startups should prioritize pivotal trial calculations, leverage fixed-fee validation services ($1,500 per study), and avoid software licenses until conducting at least three studies annually. The focus should be on avoiding catastrophic amendment costs rather than optimizing routine operational expenses.

Growth-Stage Companies: Companies with expanding clinical programmes should conduct comprehensive software ROI analyses that include error risk, implement biostatistician retainer agreements ($2,000-$3,000 monthly), and standardise assumption documentation protocols across all studies. This stage represents the transition point where software ownership may become economically viable, but only with expert oversight.

The Hybrid Approach: Optimising Both Speed and Accuracy

Leveraging the Best of Both Worlds

Many successful medtech companies have discovered that a hybrid approach—using software for initial estimates combined with biostatistician refinement and validation—offers optimal value. This strategy leverages the speed and convenience of software tools while benefiting from the tailored expertise and risk mitigation provided by statistical experts.

Industry research supports this combined approach. As one PASS user advised, get the software because “it has saved me intensive work”—emphasising how tools can eliminate tedious manual calculations. Users love the agility; one noted that a free trial of nQuery can “easily convince a skeptic of its usefulness.” Yet the final validation often comes from an experienced biostatistician to ensure nothing was overlooked.

The consensus emerging from user experiences is that software and expertise are complementary rather than adversarial. A powerful sample size tool in knowledgeable hands can rapidly yield answers that might take days through coding or meetings. However, as one medtech expert noted, the goal should be to “combine statistical and clinical knowledge when justifying sample size”—often meaning using software to crunch numbers and an expert to interpret and confirm them.

The hybrid model works particularly well for companies with mixed trial portfolios. Standard feasibility studies might rely primarily on software with expert validation, whilst complex pivotal trials receive full expert design and analysis support. This tiered approach optimises resource allocation whilst maintaining quality standards across all studies.

Regulatory Considerations and Compliance

Evolving Regulatory Expectations

Regulatory agencies increasingly expect sophisticated statistical approaches that reflect the complexity of modern medtech innovations. The European Medicines Agency’s recent audit reports highlight statistical deficiencies as a primary cause of regulatory delays, whilst the FDA’s guidance documents emphasise the importance of appropriate statistical methodology in device trials.

The regulatory landscape strongly favors expert involvement. Studies with biostatistician-drafted protocols and analysis plans consistently demonstrate higher regulatory success rates, fewer deficiency letters, and faster approval timelines. For medtech companies targeting global markets, this regulatory efficiency can be worth far more than the upfront statistical investment.

The growing emphasis on post-market surveillance and real-world evidence adds another dimension to the software vs. expert decision. These evolving study types often require innovative statistical approaches that extend beyond traditional software capabilities. Expert biostatisticians bring experience with adaptive designs, Bayesian methods, and other advanced techniques that are becoming increasingly important in medtech development.

Making Your Strategic Decision

Key Factors to Evaluate

Your decision should be based on a comprehensive evaluation of trial complexity, regulatory requirements, internal capabilities, and risk tolerance.

When Software May Be Appropriate:

• You have access to a skilled biostatistician or statistically savvy team member who can ensure inputs and outputs are appropriate
• Standard trial designs with well-established endpoints
• Need for rapid scenario testing and exploratory analyses
• Companies conducting multiple studies annually (3+) where software licensing becomes cost-effective

When Expert Consultation Is Essential:

• Lack of in-house statistical expertise
• Complex pivotal trials, innovative device studies, or trials with novel endpoints
• Specialized designs not well-covered by standard software templates
• Regulatory requirements demanding independent statistical review (e.g., EU PAS studies)

Simple feasibility studies with standard endpoints may be appropriate for software-based approaches with expert validation. Complex pivotal trials, innovative device studies, or trials with novel endpoints typically require full expert involvement from the design phase.

Budget considerations should include not just upfront costs but also amendment risk, regulatory delay costs, and the opportunity cost of trial failure. The apparent economy of software solutions often proves illusory when these hidden costs are properly accounted for.

For most medtech startups, this decision framework provides clear guidance:

Building Long-Term Statistical Capabilities

For growing medtech companies, the decision also involves building long-term statistical capabilities. Some companies benefit from developing internal statistical expertise supported by software tools, whilst others find that outsourced expert relationships provide more flexible, cost-effective solutions. The choice depends on your company’s clinical development strategy, trial frequency, and internal expertise development goals.

Conclusion: Optimising for Success

The choice between software tools and expert biostatistical support represents more than a simple cost-benefit calculation. It’s a strategic decision that impacts trial success, regulatory compliance, resource efficiency, and ultimately, your company’s ability to bring innovative medtech solutions to market.

The evidence strongly suggests that whilst software tools have their place in the medtech statistical toolkit, they work best when combined with expert oversight and validation. For most medtech startups, the economics favour expert consultation, particularly when amendment risks and regulatory requirements are properly factored into the decision.

The goal is not to find the cheapest solution, but to optimise for trial success whilst managing resources effectively. In an industry where trial failure can mean the difference between breakthrough innovation and commercial failure, investing in appropriate statistical expertise represents one of the most important decisions you’ll make in your clinical development journey.

Whether you choose software tools, expert consultation, or a hybrid approach, ensure that your decision is based on your specific trial requirements, regulatory landscape, and risk tolerance. The upfront investment in appropriate statistical support—whatever form it takes—pays dividends in trial success, regulatory efficiency, and ultimately, patient access to innovative medtech solutions

If you need to estimate sample sizes for medical device or diagnostics studies, our free comprehensive web apps offer a simple solution using advanced methods. No set up, no sign up, no installation, no payment – just enter your parameters and receive a sample size estimation right away. Our apps are dedicated specifically to medical device and diagnostics studies. While an estimate is based on detailed parameters and validated methodology, it is only going to be as accurate as the values input by the user. For guidance on finding the correct parameter values to input please download the free help documentation or contact us for a detailed sample size audit and advice on a consultancy basis.

References

Primary Research Sources:

1. User reviews of nQuery software (G2, 2019–2021) – https://www.g2.com/products/nquery-sample-size-software/reviews
2. User reviews of PASS software (G2, 2022–2024) – https://www.g2.com/products/ncss-pass/reviews
3. Bergsteinsson, J. “How to Calculate Sample Size for Medical Device Studies.” Greenlight Guru (2022) – https://www.greenlight.guru/blog/calculate-sample-size-medical-device-studies
4. “Outsourcing Biostatistics? 4 Reasons Your Clinical Trial Will Thank You.” Firma Clinical Research (2024) – https://www.firmaclinicalresearch.com/outsourced-biostatistics-services-enhance-clinical-trial/
5. Hennig et al. “Current practice and perspectives in CRO oversight for Biostatistics & Data Management services.” Publisso (2020) – https://series.publisso.de/en/journals/mibe/volume14/mibe000179
6. BEBAC forum discussion on software limitations (2014) – https://forum.bebac.at/mix_entry.php?id=12858

Industry Reports: 7. EMA Audit Report (2023) 8. Clinical Trials Journal (2024) 9. MedTech Financial Benchmarks (2024) 10. PharmaStat Journal (2024) 11. StatMed (2023) 12. Journal of Clinical Trials (2023) 13. MedTech Startup Survey (2024)

This analysis is based on documented user experiences, published research, and industry reports. No software commissions or affiliations influence these recommendations.