The Good Pharma Scorecard is a vital tool in ensuring ethical and responsible data-sharing, say Jennifer E. Miller and Amy Price.

There are many questions over who should have access to health data and who owns them.  These issues are complicated in clinical trials where data are co-produced by participants, researchers, clinicians, funders, and industry. Ninety percent of all data ever produced were produced in the past two years, raising novel ethical questions about the responsibilities of data collectors and the rights of participants. [1] These questions are particularly acute for clinical trial data, which hold life-saving potential and can advance patient and population health. 

While many advocate for enhanced transparency and agency around how health data from electronic health records, wearables, [2] and web searches (70-80% of adults search the internet with their health-related questions) are used, this hasn’t been the dominant approach for clinical trial data. Most stakeholders, from patients, [3] to funders, agree that trial data should be broadly accessible for secondary research purposes, with some even arguing that they are a public good. [4] While expectations around clinical trial data have moved towards greater levels of openness and a sense that data sharing should be a routine part of clinical research, trial data can remain hidden. [5] 

We owe it to patients participating in research to make the data they help generate widely and responsibly available. [6] The majority desire data sharing, viewing it as a natural extension of their commitment to advance scientific research. [3] Data sharing in clinical research is critical to help encourage a host of public health benefits, including facilitating independent re-analysis of data and an improved understanding of a medicine’s safety and efficacy. 

Data sharing can also accelerate innovation and the development of new cures and therapies, as scientists learn and build on each other’s work. Data sharing during the 2014 Ebola virus outbreak in West Africa, for example, helped scientists rapidly trace the origins of the virus and help control the epidemic. [7,8] Research duplication can also be reduced when resources are pooled through data sharing. For instance, sharing placebo or comparator arm trial data can decrease the number of trial participants needed for future trials of similar compounds. Placebo and comparator arm data have been successfully pooled for use in other trials from the Datasphere Project, the Coalition Against Major Diseases, and ePlacebo. [9]

To resolve this challenge, researchers from Yale University, Stanford University, and Bioethics International, supported by Arnold Ventures, created and validated measures to define good data sharing practices and to benchmark data-sharing practices by pharmaceutical companies. The researchers partnered with key stakeholders including the public and end-users and they collaborated to build the Good Pharma Scorecard (GPS), an annual ranking of new FDA approved drugs and their pharmaceutical company sponsors on their ethics performance, including clinical trial transparency and data-sharing criteria. Ranking and rating companies have long been an evidence-based method for improving corporate behaviors on critical social responsibility goals. [10,11]

A new research paper has found that the Good Pharma Scorecard rankings have improved data-sharing by pharmaceutical companies, as almost half of low scoring companies improved their data-sharing scores within 30 days of receiving their low score. AstraZeneca, for example, newly committed to reporting the number of data requests it receives annually and how each request is handled (i.e. granted or rejected). Novartis committed to sharing trial data by 6 months post approval of a drug or 18 months after the trial is over, when previously it had no specified timelines for making data available. [12] Industry trial registration and results reporting practices are also improving year after year, according to the Good Pharma Scorecard measures. The median proportion of trials in patients with publicly available results at 12 months after FDA approval rose from 87% for 2012 FDA approved drugs to 100% for 2015 drugs.  

It is critical to improve drug companies’ data sharing practices, as industry sponsors 90% of the clinical research for investigational drugs and devices. [13] The GPS data sharing measures require that companies register all applicable trials and have a policy that provides access to datasets which are ready to be analysed, and Clinical Study Reports (CSR). The constitutive elements of a CSR include the statistical analysis plan, study protocol, dataset codebook, and CSR synopsis. The GPS requires companies to clearly state how data may be requested and shared no later than 6 months post FDA or EMA approval of a drug or 18 months after the trial’s completion date. GPS measures require annual reporting on data requests received by companies and whether requests are granted or rejected. 

The Good Pharma Scorecard can serve as an agent for change. Patients/carers, clinicians, researchers, policy makers and investors, can require low-scoring companies to commit to reform before working with them. Advocacy groups might demand that companies fix any transparency issues highlighted in the GPS, as a condition for collaborating on participant trial recruitment. Formularies and prescription guideline writers could report the number of trials a company conducts to gain regulatory approval of a drug, and the proportion of those trials that are published in the medical literature.  

Empowering stakeholders to be effective at instigating change is one reason the GPS report the scores at both drug and company levels. The GPS provides a unique platform to discuss and address critical ethical challenges raised through healthcare innovation. Given several years of demonstrated impact, it is well poised to tackle other important issues, like drug pricing.

By sharing patient level trial data with the whole research community, we can catalyze the benefits of the resources, time and effort devoted to clinical trial research and advance patient and population health. [14] The Good Pharma Scorecard aims to provide companies with a consistent, fair, and achievable set of measures, while tracking further progress toward routine data sharing in clinical research.

 

Jennifer E. Miller, Assistant Professor, Yale University School of Medicine; Founder, Bioethics International.

Amy Price, Patient Editor, Research and Evaluation The BMJ; Senior Research Advisor, Medicine X, Stanford University School of Medicine.

 

References

1. Marr, B., How Much Data Do We Create Every Day? The Mind-Blowing Stats Everyone Should Read, Forbes, 2018 https://www.forbes.com/sites/bernardmarr/2018/05/21/how-much-data-do-we-create-every-day-the-mind-blowing-stats-everyone-should-read/#2b941bc360ba 
2. Na L, Yang C, Lo C, Zhao F, Fukuoka Y, Aswani A. Feasibility of Reidentifying Individuals in Large National Physical Activity Data Sets From Which Protected Health Information Has Been Removed With Use of Machine Learning. JAMA Netw Open. Published online December 21, 20181(8):e186040. doi:10.1001/jamanetworkopen.2018.6040
3. Mello, M. M., Lieou, V. & Goodman, S. N. Clinical trial participants’ views of the risks and benefits of data sharing. N. Engl. J. Med. 378, 2202–2211 (2018)
4. Rodwin MA, Abramson JD. Clinical Trial Data as a Public Good. JAMA. 2012;308(9):871–872. doi:10.1001/jama.2012.9661
5. Loder Elizabeth. Sharing data from clinical trials: Where we are and what lies ahead BMJ  2013; 347:f4794
6. Jeffrey M. Drazen, MD, Drazen JM: Data sharing and the journal. N Engl J Med 374:e24, 2016
7. Kiley R, PeatfieldT, Hansen J, Reddington, F. Data sharing from clinical trials—a research funder’s perspective. N Engl J Med 2017;377:1990–1992.
8. Armando Arias, Simon J. Watson, Danny Asogun, et al, Rapid outbreak sequencing of Ebola virus in Sierra Leone identifies transmission chains linked to sporadic cases, Virus Evolution, Volume 2, Issue 1, January 2016, vew016, https://doi.org/10.1093/ve/vew016
9. Olson, S and Downet, AS., Sharing Clinical Research Data: Workshop Summary, The National Academies Press, 2013, p. 75
10. Miller, JE. From Bad Pharma to Good Pharma: Aligning Market Forces with Good and Trustworthy Practices through Accreditation, Certification, and Rating. The Journal of Law, Medicine & Ethics, 2013, 41(3); 601-610
11. Chatterji, AK and Toffel, M.W. How firms respond to being rated. Strategic Management Journal, 2010, 31(9); 917-945.
12. Miller, JE., Ross, JS, Wilenzick, M, and Mello, MM, Sharing of clinical trial data and results reporting practices among large pharmaceutical companies: cross sectional descriptive study and pilot of a tool to improve company practices, BMJ 2019;366:l4127 http://dx.doi.org/10.1136/bmj.l4127
13. Getz, K. Sizing up the Clinical Research Market, Applied Clinical Trials, 2010 http://www.appliedclinicaltrialsonline.com/print/213683?page=full
14. Ross, JS,  Lehman, R, and  Gross, CP. The Importance of Clinical Trial Data Sharing: Towards more open science, Cardiovascular Quality and Outcomes. 2012;5:238–240