PIPEDREAM: Pharmaceutical Innovation, Progression Errors, Development Risks, Efficiency and Attrition in Medicine

PIPEDREAM: Pharmaceutical Innovation, Progression Errors, Development Risks, Efficiency and Attrition in Medicine | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Method Article PIPEDREAM: Pharmaceutical Innovation, Progression Errors, Development Risks, Efficiency and Attrition in Medicine Dennis Lendrem, Clare Lendrem This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6198126/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The PIPEDREAM model estimates false positive and false negative rates for each phase of the drug development process using data on attrition and marketing rates for each phase of development. We illustrate the model using aggregated historical data for the pharmaceutical industry from 1997-2007. These data suggest that false positive rates were high, and false negative rates low, throughout the early phases of development during this period. False positive rates were lowest and false negative rates peaked during Phase 2 clinical development. Furthermore, there was evidence of progression-seeking behaviour in Phase 3, with more molecules progressing to submission than might be expected. Translational Medicine Strategy Drug Development Costs Risks False Positives False Negatives False Discoveries Opportunity Costs Decision Making PIPEDREAM Figures Figure 1 Figure 2 INTRODUCTION At each stage of pharmaceutical development, decision makers seek to advance safe and effective medicines and terminate ineffective medicines (or those with an unacceptable safety profile) as quickly and inexpensively as possible. This gives rise to two kinds of decision error. These are false positives, or false discoveries, where ineffective drugs are progressed only to be terminated later in development; and false negatives, or missed opportunities, in which a potentially safe and effective drug is terminated 1 – 4 . In this paper we present a simple method of estimating false positive and false negative rates for each stage of the drug development process using data on progression rates and marketing rates from the pharmaceutical industry. While this estimation method is most likely to be useful in modelling in-house research and development processes, we illustrate the method using aggregated historical pharmaceutical industry data 2 , 5 . METHODS Model Overview The PIPEDREAM model is a decision-analytic framework designed to estimate false positive ( FP ) and false negative ( FN ) rates at each phase of drug development. Let p be the probability that a drug is marketable – that it is safe and effective with a clear clinical need, an acceptable safety profile, the therapeutic effect of interest and a market allowing a return on investment. Let FP be the probability of advancing a drug given that the drug is ineffective (a False Positive), and FN be the probability of failing to advance an effective drug given that the drug is marketable (a False Negative). The proportion of drugs advanced to the next stage in development is given by pA = p (1- FN ) + (1- p ) FP. Of these drugs, a proportion pT = p (1- FN ) / pA are true positives. The remainder pF = 1- pT are false positives. True positives will make it to market iff they survive False Negative decisions and are not falsely terminated during subsequent phases of the drug development process. Thus the probability of a marketable drug making it to market is given by p x (1- FN i=1 ) x (1- FN i=2 ) x (1- FN i=3 ) x….x (1- FN i=n ) where FN i are the false negative rates for each phase of development, i , and n is the number of phases in development. Now, while FP and FN at each stage may be unknown, industry estimates exist both for the percentage of candidates advanced at each phase in development, pA i , and the percentage subsequently making it to market, pM i . This means that for any starting probability, p , we can estimate FP i and FN i numerically by minimizing the deviation of modelled and industry estimates of pA i and pM i . This permits estimation of R&D Efficiency for the pharmaceutical development process defined as the probability that a marketable drug survives the R&D development process to be marketed (see Appendix: The PIPEDREAM Model ). The Data For illustration purposes, we used published industry data from the Pharmaceutical Benchmarking Forum for the years 1997–2007 2,5 . We used the percentage of molecules progressing to the next phase to estimate pA i , and the percentage at each phase that went on to be marketed to estimate pM i - see Table 1 . Table 1 Industry data for the proportion of molecules progressing to the next phase, and the proportion marketed given completion of each phase of the research and development process. 2,5 Observed Probabilities Development Phase i = 1 TTH i = 2 HTL i = 3 LO i = 4 PC i = 5 P1 i = 6 P2 i = 7 P3 i = 8 STL pAdvance, pA i 0.800 0.750 0.850 0.690 0.540 0.340 0.700 0.910 pMarketed, pM i 0.0412 0.0514 0.0686 0.0807 0.1170 0.2166 0.6370 0.9100 Key: Target to Hit (TTH), Hit to Lead (HTL), Lead Optimization (LO), Preclinical (PC), Phase 1 (P1), Phase 2 (P2), Phase 3 (P3), and Submission to Launch (STL). RESULTS False Negative and False Positives Estimates of False Negative and False Positive rates were calculated using the PIPEDREAM model for each phase of the development process and for initial starting probabilities, p(Marketable), of 5% through to 15%, and for low, modest and high correlations between phases from ρ = 0.3 through to ρ = 0.7 . The results are shown in Fig. 1. R-code is included permitting users to explore other choices for these parameters matching their own development data. See Supplementary Data . To permit a more detailed analysis of False Positives and False Negatives at each phase of development, estimates with uncertainty intervals are presented for a representative example using an intermediate value for the starting probability, p(Marketable) = 10%, and modest correlations ρ = 0.5 , as presented in Fig. 2 . The general patterns for False Negatives and False Positives are consistent for all values of the starting probability that a molecule is marketable, p , from 5%-15%. As might be expected, False Positive rates are higher for the initial phases of the development process, dropping to less than 50% during clinical trials. False Positive rates are lowest – less than 20% - for Phase 2 trials. False Negative rates are more sensitive to the starting probability, p(Marketable) . As p(Marketable) increases, the pipeline is more enriched and False Negative rates must increase to account for the observed attrition and losses in marketable molecules during drug development. However, regardless of the initial starting probability, False Negative rates are relatively low until Phase 1 trials. By Phase 2, False Negative rates exceed False Positive rates for the first time in the development process. R&D Efficiency One measure of R&D Efficiency is the probability that a marketable drug is marketed. We can estimate R&D Efficiency as a function of the FN rates for each phase – see Appendix: The Pipedream Model . Estimates of False Positive and False Negative rates together with lower and upper 95% empirical uncertainty limits are given in Table 2 below. Table 2 – Estimates (with 95% empirical intervals) for False Positive and False Negative Rates for each phase of the pharmaceutical development process based upon aggregated historical data from the Pharmaceutical Benchmarking Forum. 2,5 Development i False Negatives False Positives Phase Median Lower Upper Median Lower Upper TTH 1 0.105 0.016 0.188 0.782 0.772 0.791 HTL 2 0.088 0.014 0.177 0.719 0.710 0.729 LO 3 0.066 0.015 0.142 0.831 0.821 0.846 PC 4 0.080 0.038 0.149 0.643 0.634 0.652 P1 5 0.136 0.054 0.206 0.461 0.449 0.476 P2 6 0.162 0.103 0.205 0.093 0.086 0.109 P3 7 0.233 0.178 0.250 0.465 0.432 0.483 Key: Target to Hit (TTH), Hit to Lead (HTL), Lead Optimization (LO), Preclinical (PC), Phase 1 (P1), Phase 2 (P2), Phase 3 (P3), and Submission to Launch (STL). Assuming baseline marketability of 10% and modest correlations between phases (ρ = 0.5). The FP and FN rates for the final phase, STL, are fixed at 0.090 and zero, respectively. See text for details. Assuming a FN rate of zero for the final phase, we can use estimated FN rates for each phase to calculate the overall efficiency of the R&D process as: Under the model assumptions, less than half of all marketable molecules are likely to survive the research and development process. In addition, we can estimate the efficiency of components of the R&D process. For example, of the marketable molecules entering phase 1 we can expect to lose 13.6% by the end of Phase 1, 16.2% during Phase 2, and 23.3% of marketable molecules by the end of Phase 3. Thus, the efficiency of the clinical process is given by: Under the model assumptions, a little over half of all marketable molecules are likely to survive clinical development and make it to the marketplace. DISCUSSION The PIPEDREAM model estimates both FP and FN at each phase of the pharmaceutical development process, providing a picture of decision errors during drug development. The model can be adapted to different therapeutic areas or development pipelines by adjusting observed pA and pM rates. In the worked example, drawing upon aggregated pharmaceutical industry data from the Pharmaceutical Benchmarking Forum 2 , 5 , the false positive rates are high, and the false negatives low, throughout the early preclinical phases. Interestingly, false positive rates are lowest and false negative rates highest during Phase 2 clinical development. This may reflect an understandable reluctance on the part of decision makers to enter the most expensive phase of clinical development with equivocal Phase 2 studies. In contrast, false positive rates during Phase 3 are higher. This is consistent with strong progression-seeking behaviour in Phase 3 with pharmaceutical companies preferring to risk rejection and proceed to submission and product launch, rather than terminate at so late a stage in development. 4 , 9 The PIPEDREAM model allows us to: • Identify Bottlenecks: Highlight phases with high FN rates where potentially marketable molecules are being lost. • Optimize Decision-Making: Provide insights into how changes in progression criteria could reduce FP and FN rates. • Benchmark Performance: Compare the efficiency of different development pipelines or therapeutic areas. Of course, there are limitations both to the data and to the PIPEDREAM model itself. The estimated FP and FN rates are based upon historical data derived from pharmaceutical companies participating in the Pharmaceutical Benchmarking Forum aggregated across all therapeutic areas. More recent data on progression and marketing rates across all phases of discovery and development may not be available, but the model is flexible and permits estimation of decision risks for the clinical phases of the development process for which more recent data are available for individual therapeutic areas. The PIPEDREAM model largely sidesteps the issue of independence of the phases by modelling the probabilities of market launch at each phase conditional upon progression to that phase. However, the PIPEDREAM model incorporates potential dependencies by including an autoregressive correlation structure with low, moderate, and high dependencies between phases. The choice of a stationary correlation structure where the pattern of dependencies between phases remains constant throughout the entire development pipeline. The AR1 model assumes that the correlation between Phase 1 & 2 FP rates and the correlation between Phase 6 & 7 FP rates are both equal with ρ^1 = ρ , despite being different phase pairs. This does not rule out carryover between phases. For example, we might expect that a drug that sails through, say, Phase 1 is more likely to progress through to Phase 2, than a drug which narrowly makes it through Phase 1. In this case the model would underestimate progression rates in Phase 2. Conversely, the Winner’s Curse might mean that a drug that sails through Phase 2 with an inflated overestimate of effect size is less likely to make it through Phase 3 as the model will overestimate Phase 3 progression rates. In addition, real-world development processes may exhibit phase-dependent decay with later phases (e.g., P3 to STL) showing stronger correlation persistence than early phases (TTH to HTL). In addition, there may be regime changes at key points. Major phase transitions (e.g., preclinical to clinical) could alter dependence patterns. In addition, external influences such as changes in trial designs over the last ten years or shifting regulatory landscapes might modify correlations over time. While a stationary correlation structure enables tractable modelling, it may oversimplify complex phase transition dynamics, underestimate uncertainty in later phases, and might best be improved with phase-specific correlation data. Furthermore, the model assumes that the final, STL, phase is an absorptive state with known FP and FN rates where FP STL = 1 - pM STL and FN = 0. However, this is not critical. The model allows users to experiment with other choices for FP STL and FN STL . Nevertheless, this analysis gives some insight into recent drug development strategies. In particular, the pharmaceutical development pipeline may be richer in marketable drugs than currently supposed. During the period through to 2010, just 4% of projects entering development survived the pharmaceutical development process and made it to market. Many of these marketable drugs may be lost during Phase 2 trials. This analysis supports the strategic targeting of Phase 2 trials and the need to better power Phase 2 studies to reduce false negatives. 2 , 3 , 6 At least for the clinical phases, more recent data 18 suggest that pA and pM have changed little since the Pharmaceutical Benchmarking Forum of 2010. We welcome more recent data for all the phases of the current pharmaceutical research and development process. References Kola, I. & Landis, J. Can the pharmaceutical industry reduce attrition rates? Nat Rev Drug Discov 3, 711-715, doi:10.1038/nrd1470 (2004) Paul, S. M. et al. How to improve R&D productivity: the pharmaceutical industry's grand challenge. Nat Rev Drug Discov 9, 203-214, doi:10.1038/nrd3078 (2010). Lindborg, S. R., Persinger, C. C., Sashegyi, A., Mallinckrodt, C. & Ruberg, S. J. Statistical refocusing in the design of Phase II trials offers promise of increased R&D productivity. Nat Rev Drug Discov 13, 638-640, doi:10.1038/nrd3681-c1 (2014). Lendrem, D. W. et al. Progression-seeking bias and rational optimism in research and development. Nat Rev Drug Discov 14, 219-221, doi:10.1038/nrd4320-c1 (2015). Dimitri, N. An assessment of R&D productivity in the pharmaceutical industry. Trends in pharmacological sciences 32, 683-685, doi:10.1016/j.tips. 2011.09.005 (2011). Owens, P. K. et al. A decade of innovation in pharmaceutical R&D: the Chorus model. Nat Rev Drug Discov 14, 17-28 doi:10.1038/nrd4497 (2015). Lendrem, D. W. & Lendrem, B. C. Torching the Haystack: modelling fast-fail strategies in drug development. Drug Discov. Today 18, 331-336 (2013). Peck, R. W. Driving earlier clinical attrition: if you want to find the needle, burn down the haystack. Considerations for biomarker development. Drug Discov. Today 12, 289-294 (2007). Peck, R.W. et al. Why is it hard to terminate failing projects in pharmaceutical R&D? Nat Rev Drug Discov 14, 1-2, doi10.1038/nrd4725 (2015) Scrucca L “GA: A Package for Genetic Algorithms in R.” Journal of Statistical Software, 53(4), 1-37. doi:10.18637/jss.v053.i04 https://doi.org/10.18637/jss.v053.i04 (2013). Scrucca L “On some extensions to GA package: hybrid optimisation, parallelisation and islands evolution.” The R Journal, 9(1), 187-206. doi:10.32614/RJ-2017-008 https://doi.org/10.32614/RJ-2017-008 (2017). Marius Hofert, Ivan Kojadinovic, Martin Maechler and Jun Yan “copula: Multivariate Dependence with Copulas. R package version 1.1-5” https://CRAN.R-project.org/package=copula (2025). Jun Yan Enjoy the Joy of Copulas: With a Package copula. Journal of Statistical Software, 21(4), 1-21. https://www.jstatsoft.org/v21/i04/ (2007). Marius Hofert, Martin Maechler Nested Archimedean Copulas Meet R: The nacopula Package. Journal of Statistical Software, 39(9), 1-20. URL https://www.jstatsoft.org/v39/i09/ (2011) Corporation M, Weston S doParallel: Foreach Parallel Adaptor for the 'parallel' Package_. R package version 1.0.17, https://CRAN.R-project.org/package=doParallel (2022). Wickham H, François R, Henry L, Müller K, Vaughan D dplyr: A Grammar of Data Manipulation_. R package version 1.1.4, https://CRAN.R-project.org/package=dplyr (2023). R Core Team R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/ R version 4.4.2 (2024-10-31 ucrt) Knowledge Portal on Innovation and Access to Medicines, Research Synthesis: Time and Success Rates of Pharmaceutical R&D. https://www.knowledgeportalia.org/r-d-time-and-success-rate (2020) Additional Declarations The authors declare potential competing interests as follows: No competing interests to declare. Supplementary Files APPENDIX.docx SUPPLEMENTARYDATA.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6198126","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Method Article","associatedPublications":[],"authors":[{"id":426882488,"identity":"a2c83d5c-3d5f-48ea-be82-e2d27ab3ccf6","order_by":0,"name":"Dennis Lendrem","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABAklEQVRIiWNgGAWjYDCCA0CcwMAgA6QMgNiGgYEdImFASAsPVFEaAwMzMVoYEFoOE9bCd7z32IMHfxh4+KWbt334mHM+j7+ZgfHDD4bDxri0SJ45l26Q2MbAIznnWPHMmdtuF0scZmCW7GE4bIZLi8GNHDOJxAYGHiDDmJl32+3EDUCHSQNdaINTy/03ZhIJQIfZg7T83XYOpIX5N14tN3iAWtiAtkgAtTBuOwDSwgayBafDJM+AHNYmwSNxI62YsXdbcuKMw4xtlj0G6Ti9z3f8jJnkjz82cvwzkjcz/Nxml9jf3nz4xo8Ka8MGXHogQAKZw9iALyJHwSgYBaNgFBABAD9hUARUaowMAAAAAElFTkSuQmCC","orcid":"","institution":"Newcastle University Translational \u0026 Clinical Research Institute","correspondingAuthor":true,"prefix":"","firstName":"Dennis","middleName":"","lastName":"Lendrem","suffix":""},{"id":426882644,"identity":"6f5aaae9-649d-436d-a379-3351f67e0b65","order_by":1,"name":"Clare Lendrem","email":"","orcid":"","institution":"Newcastle University Translational \u0026 Clinical Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Clare","middleName":"","lastName":"Lendrem","suffix":""}],"badges":[],"createdAt":"2025-03-10 19:26:07","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":true,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-6198126/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6198126/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":87844357,"identity":"8e89b7de-759e-4076-a60f-83b716cc18d4","added_by":"auto","created_at":"2025-07-29 14:44:58","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":573774,"visible":true,"origin":"","legend":"\u003cp\u003eEstimates (Medians and 95% Empirical Uncertainty Intervals) for False Negatives (\u003cem\u003eFN\u003c/em\u003e) and False Positives (\u003cem\u003eFP\u003c/em\u003e) for values of \u003cem\u003ep(Marketable)\u003c/em\u003e from 5% through to 15% and A) low, B medium and C) high phase dependencies (\u003cem\u003eρ=\u003c/em\u003e0.3, \u003cem\u003eρ=\u003c/em\u003e0.5 and \u003cem\u003eρ=\u003c/em\u003e0.7). See text for details.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6198126/v1/d49a3dd421085258253eb32a.png"},{"id":87843511,"identity":"698af937-df69-4e40-8702-52849f4c7031","added_by":"auto","created_at":"2025-07-29 14:36:59","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":72215,"visible":true,"origin":"","legend":"\u003cp\u003eEstimates (medians and 95% empirical uncertainty intervals) of False Negatives (FN in \u003cstrong\u003ered\u003c/strong\u003e) and False Positives (FP in \u003cstrong\u003egrey\u003c/strong\u003e) for the pharmaceutical industry given an intermediate starting probability that a molecule is marketable of 10%.\u0026nbsp; Worked example using historical, aggregrated industry data, with phase dependencies \u003cem\u003eρ\u003c/em\u003e=0.5.\u0026nbsp; Note that decision errors for the Submission to Launch (STL) phase are fixed at FP\u003csub\u003eSTL\u003c/sub\u003e = 1 – pM\u003csub\u003eSTL\u003c/sub\u003e = 1 – 0.910 = 0.090 and FN\u003csub\u003eSTL\u003c/sub\u003e is fixed at zero.\u0026nbsp; See text for details.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6198126/v1/d88bd8f54ce8b0c9292127d2.png"},{"id":87844846,"identity":"24571f15-a4ac-4455-8941-35bf40b3196c","added_by":"auto","created_at":"2025-07-29 14:52:59","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1243509,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6198126/v1/0e82fcf9-d058-47d9-b0b6-a3175eb49480.pdf"},{"id":87843507,"identity":"061af288-1469-43c9-a57b-11ae2271af9b","added_by":"auto","created_at":"2025-07-29 14:36:58","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":23323,"visible":true,"origin":"","legend":"","description":"","filename":"APPENDIX.docx","url":"https://assets-eu.researchsquare.com/files/rs-6198126/v1/95ac942cca7bd8946eaa11bb.docx"},{"id":87843509,"identity":"41754e08-ec5e-497c-b161-a872fb1fc3f1","added_by":"auto","created_at":"2025-07-29 14:36:58","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":23901,"visible":true,"origin":"","legend":"","description":"","filename":"SUPPLEMENTARYDATA.docx","url":"https://assets-eu.researchsquare.com/files/rs-6198126/v1/27573b5b1611436432332c8d.docx"}],"financialInterests":"The authors declare potential competing interests as follows: No competing interests to declare.","formattedTitle":"\u003cp\u003ePIPEDREAM: Pharmaceutical Innovation, Progression Errors, Development Risks, Efficiency and Attrition in Medicine\u003c/p\u003e","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eAt each stage of pharmaceutical development, decision makers seek to advance safe and effective medicines and terminate ineffective medicines (or those with an unacceptable safety profile) as quickly and inexpensively as possible. This gives rise to two kinds of decision error. These are false positives, or false discoveries, where ineffective drugs are progressed only to be terminated later in development; and false negatives, or missed opportunities, in which a potentially safe and effective drug is terminated \u003csup\u003e\u003cspan additionalcitationids=\"CR2 CR3\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e–\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eIn this paper we present a simple method of estimating false positive and false negative rates for each stage of the drug development process using data on progression rates and marketing rates from the pharmaceutical industry. While this estimation method is most likely to be useful in modelling in-house research and development processes, we illustrate the method using aggregated historical pharmaceutical industry data \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e"},{"header":"METHODS","content":"\u003cp\u003eModel Overview\u003c/p\u003e\u003cp\u003eThe PIPEDREAM model is a decision-analytic framework designed to estimate false positive (\u003cem\u003eFP\u003c/em\u003e) and false negative (\u003cem\u003eFN\u003c/em\u003e) rates at each phase of drug development.\u003c/p\u003e\u003cp\u003eLet \u003cem\u003ep\u003c/em\u003e be the probability that a drug is marketable – that it is safe and effective with a clear clinical need, an acceptable safety profile, the therapeutic effect of interest and a market allowing a return on investment. Let \u003cem\u003eFP\u003c/em\u003e be the probability of advancing a drug given that the drug is ineffective (a False Positive), and \u003cem\u003eFN\u003c/em\u003e be the probability of failing to advance an effective drug given that the drug is marketable (a False Negative). The proportion of drugs advanced to the next stage in development is given by \u003cem\u003epA\u003c/em\u003e = \u003cem\u003ep\u003c/em\u003e (1-\u003cem\u003eFN\u003c/em\u003e) + (1-\u003cem\u003ep\u003c/em\u003e) \u003cem\u003eFP.\u003c/em\u003e Of these drugs, a proportion \u003cem\u003epT\u003c/em\u003e = \u003cem\u003ep\u003c/em\u003e (1-\u003cem\u003eFN\u003c/em\u003e) / \u003cem\u003epA\u003c/em\u003e are true positives. The remainder \u003cem\u003epF\u003c/em\u003e = 1- \u003cem\u003epT\u003c/em\u003e are false positives.\u003c/p\u003e\u003cp\u003eTrue positives will make it to market iff they survive False Negative decisions and are not falsely terminated during subsequent phases of the drug development process. Thus the probability of a marketable drug making it to market is given by \u003cem\u003ep\u003c/em\u003e x (1-\u003cem\u003eFN\u003c/em\u003e\u003csub\u003e\u003cem\u003ei=1\u003c/em\u003e\u003c/sub\u003e) x (1-\u003cem\u003eFN\u003c/em\u003e\u003csub\u003e\u003cem\u003ei=2\u003c/em\u003e\u003c/sub\u003e) x (1-\u003cem\u003eFN\u003c/em\u003e\u003csub\u003e\u003cem\u003ei=3\u003c/em\u003e\u003c/sub\u003e) x….x (1-\u003cem\u003eFN\u003c/em\u003e\u003csub\u003e\u003cem\u003ei=n\u003c/em\u003e\u003c/sub\u003e) where \u003cem\u003eFN\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e are the false negative rates for each phase of development, \u003cem\u003ei\u003c/em\u003e, and \u003cem\u003en\u003c/em\u003e is the number of phases in development.\u003c/p\u003e\u003cp\u003eNow, while \u003cem\u003eFP\u003c/em\u003e and \u003cem\u003eFN\u003c/em\u003e at each stage may be unknown, industry estimates exist both for the percentage of candidates advanced at each phase in development, \u003cem\u003epA\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e, and the percentage subsequently making it to market, \u003cem\u003epM\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e. This means that for any starting probability, \u003cem\u003ep\u003c/em\u003e, we can estimate \u003cem\u003eFP\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003eFN\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e numerically by minimizing the deviation of modelled and industry estimates of \u003cem\u003epA\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003epM\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e. This permits estimation of R\u0026amp;D Efficiency for the pharmaceutical development process defined as the probability that a marketable drug survives the R\u0026amp;D development process to be marketed (see \u003cb\u003eAppendix: The PIPEDREAM Model\u003c/b\u003e).\u003c/p\u003e\u003cp\u003eThe Data\u003c/p\u003e\u003cp\u003eFor illustration purposes, we used published industry data from the Pharmaceutical Benchmarking Forum for the years 1997–2007\u003csup\u003e2,5\u003c/sup\u003e. We used the percentage of molecules progressing to the next phase to estimate \u003cem\u003epA\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e, and the percentage at each phase that went on to be marketed to estimate \u003cem\u003epM\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e - see Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eIndustry data for the proportion of molecules progressing to the next phase, and the proportion marketed given completion of each phase of the research and development process.\u003csup\u003e2,5\u003c/sup\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"9\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cem\u003eObserved Probabilities\u003c/em\u003e\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"8\" nameend=\"c9\" namest=\"c2\"\u003e\u003cp\u003e\u003cem\u003eDevelopment Phase\u003c/em\u003e\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cem\u003ei\u003c/em\u003e = 1\u003c/p\u003e\u003cp\u003e\u003cb\u003eTTH\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cem\u003ei\u003c/em\u003e = 2\u003c/p\u003e\u003cp\u003e\u003cb\u003eHTL\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cem\u003ei\u003c/em\u003e = 3\u003c/p\u003e\u003cp\u003e\u003cb\u003eLO\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cem\u003ei\u003c/em\u003e = 4\u003c/p\u003e\u003cp\u003e\u003cb\u003ePC\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u003cem\u003ei\u003c/em\u003e = 5\u003c/p\u003e\u003cp\u003e\u003cb\u003eP1\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e\u003cem\u003ei\u003c/em\u003e = 6\u003c/p\u003e\u003cp\u003e\u003cb\u003eP2\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e\u003cem\u003ei\u003c/em\u003e = 7\u003c/p\u003e\u003cp\u003e\u003cb\u003eP3\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e\u003cem\u003ei\u003c/em\u003e = 8\u003c/p\u003e\u003cp\u003e\u003cb\u003eSTL\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cem\u003epAdvance, pA\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003e0.800\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e0.750\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cb\u003e0.850\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e0.690\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u003cb\u003e0.540\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e0.340\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e\u003cb\u003e0.700\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e\u003cb\u003e0.910\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cem\u003epMarketed, pM\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003e0.0412\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e0.0514\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cb\u003e0.0686\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e0.0807\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u003cb\u003e0.1170\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e0.2166\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e\u003cb\u003e0.6370\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e\u003cb\u003e0.9100\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"9\" nameend=\"c9\" namest=\"c1\"\u003e\u003cp\u003e\u003cem\u003eKey: Target to Hit (TTH), Hit to Lead (HTL), Lead Optimization (LO), Preclinical (PC), Phase 1 (P1), Phase 2 (P2), Phase 3 (P3), and Submission to Launch (STL).\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e"},{"header":"RESULTS","content":"\u003cp\u003eFalse Negative and False Positives\u003c/p\u003e\n\u003cp\u003eEstimates of False Negative and False Positive rates were calculated using the PIPEDREAM model for each phase of the development process and for initial starting probabilities, p(Marketable), of 5% through to 15%, and for low, modest and high correlations between phases from \u003cem\u003e\u0026rho;\u0026thinsp;=\u0026thinsp;0.3\u003c/em\u003e through to \u003cem\u003e\u0026rho;\u0026thinsp;=\u0026thinsp;0.7\u003c/em\u003e. The results are shown in Fig. 1. R-code is included permitting users to explore other choices for these parameters matching their own development data. See \u003cstrong\u003eSupplementary Data\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003eTo permit a more detailed analysis of False Positives and False Negatives at each phase of development, estimates with uncertainty intervals are presented for a representative example using an intermediate value for the starting probability, \u003cem\u003ep(Marketable)\u003c/em\u003e\u0026thinsp;=\u0026thinsp;10%, and modest correlations \u003cem\u003e\u0026rho;\u0026thinsp;=\u0026thinsp;0.5\u003c/em\u003e, as presented in Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e\n\u003cp\u003eThe general patterns for False Negatives and False Positives are consistent for all values of the starting probability that a molecule is marketable, \u003cem\u003ep\u003c/em\u003e, from 5%-15%. As might be expected, False Positive rates are higher for the initial phases of the development process, dropping to less than 50% during clinical trials. False Positive rates are lowest \u0026ndash; less than 20% - for Phase 2 trials.\u003c/p\u003e\n\u003cp\u003eFalse Negative rates are more sensitive to the starting probability, \u003cem\u003ep(Marketable)\u003c/em\u003e. As \u003cem\u003ep(Marketable)\u003c/em\u003e increases, the pipeline is more enriched and False Negative rates must increase to account for the observed attrition and losses in marketable molecules during drug development.\u003c/p\u003e\n\u003cp\u003eHowever, regardless of the initial starting probability, False Negative rates are relatively low until Phase 1 trials. By Phase 2, False Negative rates exceed False Positive rates for the first time in the development process.\u003c/p\u003e\n\u003cp\u003eR\u0026amp;D Efficiency\u003c/p\u003e\n\u003cp\u003eOne measure of R\u0026amp;D Efficiency is the probability that a marketable drug is marketed. We can estimate R\u0026amp;D Efficiency as a function of the \u003cem\u003eFN\u003c/em\u003e rates for each phase \u0026ndash; see \u003cstrong\u003eAppendix: The Pipedream Model\u003c/strong\u003e. Estimates of False Positive and False Negative rates together with lower and upper 95% empirical uncertainty limits are given in Table \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e below.\u003c/p\u003e\n\u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003e\u0026ndash; Estimates (with 95% empirical intervals) for False Positive and False Negative Rates for each phase of the pharmaceutical development process based upon aggregated historical data from the Pharmaceutical Benchmarking Forum.\u003csup\u003e2,5\u003c/sup\u003e\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eDevelopment\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" rowspan=\"2\"\u003e\n \u003cp\u003e\u003cem\u003ei\u003c/em\u003e\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colspan=\"3\"\u003e\n \u003cp\u003eFalse Negatives\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colspan=\"3\"\u003e\n \u003cp\u003eFalse Positives\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003ePhase\u003c/strong\u003e\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eMedian\u003c/strong\u003e\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eLower\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eUpper\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eMedian\u003c/strong\u003e\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eLower\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eUpper\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eTTH\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003e1\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.105\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.016\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.188\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.782\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.772\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.791\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eHTL\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003e2\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.088\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.014\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.177\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.719\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.710\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.729\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eLO\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003e3\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.066\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.015\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.142\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.831\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.821\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.846\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003ePC\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003e4\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.080\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.038\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.149\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.643\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.634\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.652\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eP1\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003e5\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.136\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.054\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.206\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.461\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.476\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eP2\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003e6\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.162\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.103\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.205\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.093\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.086\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.109\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eP3\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003e7\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.233\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.178\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.250\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.465\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.432\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.483\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colspan=\"8\"\u003e\n \u003cp\u003e\u003cem\u003eKey: Target to Hit (TTH), Hit to Lead (HTL), Lead Optimization (LO), Preclinical (PC), Phase 1 (P1), Phase 2 (P2), Phase 3 (P3), and Submission to Launch (STL). Assuming baseline marketability of 10% and modest correlations between phases (\u0026rho;\u0026thinsp;=\u0026thinsp;0.5). The FP and\u003c/em\u003e FN \u003cem\u003erates for the final phase, STL, are fixed at 0.090 and zero, respectively. See text for details.\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003c/p\u003e\n\u003cp\u003eAssuming a \u003cem\u003eFN\u003c/em\u003e rate of zero for the final phase, we can use estimated \u003cem\u003eFN\u003c/em\u003e rates for each phase to calculate the overall efficiency of the R\u0026amp;D process as:\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\"\u003e\u003c/p\u003e\n\u003cp\u003eUnder the model assumptions, less than half of all marketable molecules are likely to survive the research and development process.\u003c/p\u003e\n\u003cp\u003eIn addition, we can estimate the efficiency of components of the R\u0026amp;D process. For example, of the marketable molecules entering phase 1 we can expect to lose 13.6% by the end of Phase 1, 16.2% during Phase 2, and 23.3% of marketable molecules by the end of Phase 3.\u003c/p\u003e\n\u003cp\u003eThus, the efficiency of the clinical process is given by:\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAeEAAABDCAYAAACr6b1fAAAAAXNSR0IArs4c6QAAAARnQU1BAACxjwv8YQUAAAAJcEhZcwAADsMAAA7DAcdvqGQAABSISURBVHhe7Zzfi1fFG8fPfu/9fVcRot6IhWHaRdmFgj9SCENBiwihyLQQVFTUJIKyFCqQEAWvxPxBgiJo/oAu/BHkLxSShFQizCslyz9gv+c1e5519tk553N2P5/ts7u9X3D2nJkz55mZZ56ZZ2bO+WxHZ04mhBBCiH+d/xVnIYQQQvzLyAkLIYQQbUJOWAghhGgTcsJCCCFEm5ATFkIIIdqEnLAQQgjRJuSEhRBCiDYhJyyEEEK0iX/VCT98+DBbtmxZduHChSKmHocOHcpWrVpVhIQQQojhQaUT3rFjR9bR0VF6GOZcfTzOc9KkSSEOWffu3cvOnDmT3bx5s0hRj3PnzgVZYuBhspOa8Ny9ezcbO3ZsNn/+/CKmebzdIPvGjRvF3cbwPM/ENheDzcyYMaNbPnml5O/Zs6fbTqkjtmpYvTm3mjJdnzx5MpQlLkezUO9YF+SL/urSqP2ZWDeSTxni9iZ9PCGn3rTDQIDc1BiyZcuWUJa+LgzqEI9/3q6qaKQnQLfIi+WndM5zXhZ6Nqh/ygabBXshv1S/GYw6R291xiLKjlzSkE+sS6NRX+ulc/5tZRnbt2/nX1p27t69u/P8+fO9DoP7pDtx4kSP+DFjxnQuXbo0xN25c6eIFYMV2oh29G118ODB0JbcmzdvXhHbPNOnTw9ykY/tWPjBgwdFinKwKdJTJg6P2eTmzZtD2lh+XD+zcc6k40wanjNWrlwZjlaS0jX1tvJYmVoBeVAn6o8erD3rtmWj9kdv3LO+TvqJEyeG/AzqZmWw9kYWz12/fr1I1Rme434rsfLEoBPKa7qm3K2E+sU6ie2sirp6Qi7pbGwmrT1nkN7KgBzSWZ0JQ8oOW0GqzwxWnQN6Q3+mcwvHYxFjgsmLdRnXo05f8zqv5YQbKcvSeazAYmhAp8GwYmwAwAC5jo2pGTBOb1tmnAwsVZi9URbrGJ5UWU0+eQMdjLC3UZNvHdAGM+s0rSCla/KwgSBVrv5CXn5AsTzigT0FOiRdVfsz4Ph42pXnbLC3SVGsQ9N/POEhnXeYzYK82KasbBbPdaMxrq8g2+vE2qGKunoi7O3D2tSeNTv2ULbYQWKHcbhZrJ/FtjWYdV5nLLKw1zl5xv24bl+Ldd60E+a+P+w5H2fpfUUIUxnuUQEKF1fC5MVQISpi8lG+H1CIp+w8j1zCDBg+nZdlsxiw2ZXXgXWMuHMPddARBhND/a1To2Nv5P0FfdPmntSA7qEtrA1TtlEGz8Rt6QetKihrPAjGEO87ntlcHBeT0jVhS0+5kNEKyMsPsma/jfJo1P42OPm6eGhv2tbj29vkWf/zpOwjVS7D2jxuY65NPrbAfd+/m8HyTPWlRnnV1VMKc26NwB7idrd+kLJVdOXTk87HxVAO37cHu84bjUWm29iOgL5PvFG3r8U6r/VhFu9w2b+PD9svzyuX5Zl2X3MsWbIknIF7FpeCPfuNGzdm69evD+n2798f9vWr3j3ynmHWrFnh2vKcMGFCiPPvINasWZNdu3YtO378eJY3SJYrMeRnsFfPc8SbLPbzFy5cGPb7FyxYkOWKzQ4cOFA80cXZs2fDec6cOeE8mLD3pKmj6p3eX3/9lU2ePLmI6YL6o9tW8+jRo/BOxZN3huz06dNFKM3MmTOzqVOnFqHGUDds6r333gv2yPPwxx9/hHOd+uUdNXzPkGLt2rXhPdHXX38dwtggNvbFF19k48aNC3ExZbrmnVQqfbOQ1/jx44tQF+SDXf/4449FTJpG7X///v1wfuaZZ8K5DPR/+fLlItQFfe/KlSvZ7Nmzi5iutsgHv+zixYtFTE/27t0b7MPexSGXcD7AhbCHe9QzrgPX1GugePz4cTh7nTz99NPh/NNPP4Vzirp6MriHPfFdw6ZNm7J8cC/u9ITxGp1hY/Sxd999t7jzZAz7+eefwzkGXWHHX375ZffYunXr1mDvsYyYo0ePhjaMGcw6rzMW/f333+Ec2xE8++yz4UwbQN2+1kPnhWNOgufO0yWPeFZm6TzE+dlSHGczKr/CsNlL2WontfIAZjOxLJ5hlhPjZVnYy2IWZM9afjHc87IHC+gtn0wkD9Opp84sumrF0VfIKyXLt08j6qTnPgf2gQ4Mqw9xtLelo139jLeRfpDBffTbSE91dM196tYsVq6UrEbl9KTSm/7RF3rjmgN9xrpOYasIr2vkVJWLPGlLnqNfVumpkSzTT6Oy9gXTSUom8VXlTVGmJ7Dyc1DPVBrgnqXzq0VopEeeR5d19NVIVh0ZfaUZnXM/ZSMmE0x/nrgudp3Ki+d9HqanWivhXDi59zhOnTpV3O0/v/76a5g5vPbaa0VMF8yYyKNstXP16tXspZde6rVqYDbDvZhp06YVV2mYneTK6SWL2SizUmDGRzlt9s3s8/Dhw9miRYtCOAUzI/syj9ln6itBg1ms5VWXWL4HvbHaSx1lOmWmx8xvOIIt5ZOsbO7cudmrr77aPWuFS5cuZcuXLw+rZOyclQQrXtLGTJkypbhKg27zQSpbvHhxkLlr167iTm+Go67RF3VCh+w4Abous3vsfdu2bUHffnXRqM/aCszy3LBhQwinYJXj5Q8lqvQE2B32nTvfMC6xAuXsYbwmXT4BzN58881eYwfjaRW5swhjHn0lnxSEfMtgrLQVoijHdD5k/1kH2wR+q9W2DloNxo9x79u3L4TZis5nMbW3oulIVQPBihUrgqNuFf3ZjobUlkxf8XmX5cfEJwWvDgYCJlk4Rgbtb7/9NsSxvceAwZYQbcDAQjvwSoRBLXbWdXjnnXfCc+TRaOBvha5jPXOkJmVVgyWThWZ5+eWXw3nnzp3Z559/HvJjEm2T9CNHjoRzDP0BR4Bj6Y/d05ZMeNC1nyyl8NuDfcXr2R/eTkwnntRPXqroi56wN5wk9pzSuYGd80qG7eW+wOSdMRCdv/LKK0VsOY1eTzQipef4aKXO64xFqVcBEP/ctr99bcg6YRTHzM4frVihp1i3bl0wcmaZx44dC53CVs8YBM6Gg9+AecwRkY7nOOLf6zF4mlGxKjZZdEIgHWF+W5aS72HWyookdXBvIPF5l+U3evTo7Pbt20XoCXTysk5RF1ZfdFTTbwzOj9URjBo1Kpw9I0eOLK6e8M8//xRXabCL1atXdzsHa7uBJNYzR9l3F0wYf//99yLUBeVlwC4bXOoyYsSI4qonfmfJqONY7P1bGQysrA5xJpyrdplagdezP/y7fdMJ/xch5s8//wznMocRU6Un7mHffsVrEz/TH7aeGi+YlND2ManVcwzjEnaNfWPnA01Kz/HRSp3XGYtsrPC2Zt+VmAPuS1/r1nnuuErJB9Du/e4qLJ2HOO7FxHG8hyUcv8cF/0Wbl887NfbT/Xtc5MSyUvl7Wfa+JfVOOG+AIvSkrPY+L9ZJvvLpfp73JnnjhfuWt8khjvtAeosnHfd4jnwNu6aMJj+W5evWDF4vKcg71kkz5INLLz1Sf9OxQRxHGWXlxj5M1wY6JD4fvEPY8vN6tLLF788b6Yc2Mjsgbco+jTq6TpWrv1BfX55UHbkuKzOUtT/19vGptqQcxJF3FY3sjH5hbdsobaP72B9liu2wFaR0Yu0Qk9J5Iz3xjNctWF3sOXQUj00G5SI+hufK7I22pNzkhyyuq2yzShYMNp1bX4jL4+23bKwgz3icqdvXwORVjgQhQVEQCugPw9J5UoX2cfYsZ2TigOlksTPy8s0QqLyVhTJ62an8y2SRH3kjy9IQjjEFewP2YTA5YIYRx4HFE2f1IA8PjUi8DfTgZTUL8qgzxlYG5fVG3gzoHH1SP2t3b8CEq8plbeUxe0Bv1A35lJ24uDPYJMzagLKQp68ncsrqbrqL7cV3zpg6urYytQLyoU5m46k62sCesmWD9CkdIJNnrT+m2hL9kcbawx8xlKGs7sQj13THGbk2WHpIX1Un8uZ5X4Zm8TqhHITjeqV0XldPyEUPyCOeNkUOh+nc5NNmlId0Jj/Wl+kwlh/D87SnQV6k907FID35lEE+Vfn1l/7qHOqMRX6sIB9fjzp9DWKd13LCZYdh6TzExQqAVBxhlMI9CkwDxpVPyacSpgQOnq+bl5dFw8SyTIEea2QaIyY2fMqOPDMCsAaI48DiieMedYqN3a5TcV5WKzAjLIPyemNqBnQW651r37Gpb6xfT6o9DetMVfIBGVX2B9wvG+jJA9kxoXPl8lJ2BI10zbOtbF/qbbpI1dEGD1+PmKr2b6Rriy87DBucUu1kZfTtgJ6I574HOchL3QNrJ86thvY1u+Ls2zOlc9JWHTF17Jb6I9+ep428TZqjSGFjnm8P7ABZKWgfylXGYNM51BmLgLGf50mT0iXwXFVfg1jn6dFLJEFxKNZ3aJRuA5Q5aJROQ/CMDVwYXWwUFk+cGaSl5+AaaESTbYYfy28V5OONUzQeyPuDdJ2m0QDeH5BXNoESXatq7LFV0E/oLyknJrqIdd7Bn1xhogJext+6dSt8kMDHDgP18Ve7oZ75gBU+SGj0de9/Cftn61U/O+or0nUa+tdnn31W+tFWf+BDpo8//jj58c1/nYGyw4HoM8OFXjoPrlhUwko1111YjTLLG84wO2vlrHioQ3uzbTQQ7S5d94StPVatAwFyW7lrNFxgd20gbJD+wpg53MfL/uB1rpWwEEII0SaG7O+EhRBCiKGOnLAQQgjRJuSEhRBCiDYhJyyEEEK0CTlhIYQQok3ICQshhBBtQk5YCCGEaBNywkIIIUSbkBMWQggh2oScsBBCCNEm5ISFEEKINiEnLIQQQrQJOWEhhBCiTcgJCyGEEG1CTlgIIYRoE3LCQgghRJuQExZCCCHahJywEEII0SbkhIUQQog2IScshBBCtIl/1Qk/fPgwW7ZsWXbhwoUiph6HDh3KVq1aVYSEEEKI4UGlE96xY0fW0dFRehjmXH08znPSpEkhDln37t3Lzpw5k928ebNIUY9z584FWWJg8e04f/787MaNG8XdcsrshOdjUmk44klZXVlCCDEc6OjMKa57wYC4cePGbPfu3dmUKVOK2CfMnDkznPfs2ZN98MEH2YkTJ7KRI0d2x48dOzabO3du9tFHH2VPPfVUNmHChBAvBiczZszI7ty5k+3atSu04yeffBLCv/32WzZu3LgiVW+wk++//z775ptvipguRowYkU2dOrUIdTnhlC1Nnjy5W35dWUIIMSzACZexfft2HHTn+fPni5g0ls5DHPfE4OfgwYO92jp3wCEud5xFTJqlS5d2rly5sgiluX79epCFzCrqyBJCiOFC0++EWd2wWrZrDttSBO5ZHMTXBmHbtmb1zPtftkaNWJ7BNmmjrVPi2erkeeQSZrXn03lZpDl58mS4x9nkxFA+4tkFGA4cO3YsmzhxYvcuBrBzMX369Ozo0aNFTJpHjx5l48ePL0JpHj9+HM6NdkPqyBJCiOFCLSfMO1ycUHyYI8tXTlm+cum+5liyZEk4A/csLgUOEke9fv36kG7//v3h/W/VO8C7d+9ms2bNCteWJ4M7cdyLWbNmTXbt2rXs+PHjYbuc7VWbNADOlOeIN1k44YULFwYHvGDBgmzMmDHZgQMHiie6OHv2bDjPmTMnnAcT6M4mFP4o0yvOj4mQB8d8+vTpIlQOOkZv5JGaSBlbtmzpMSGyyU5MXVlCCDHkKVbESWybOXXMmzevSNW37eg47sGDB525g+vcvHlzCBu5swzp2MIEL5/0PMfzMbnD6CGLZ9jejPGyLOxl5SvA7mctvxjuedmDBfSWTyaSh+nUgw7iNjW8vlKgd/RDWvJga5swOjRsu5s8aF/SoT/iCBt1ZAkhxHChlhNmMKyibKAmjnsxcRxy+yOfgTzlMHx8Kv+6smLs3ag5Cxw2YRxEGdTJ8sbZVL0L5Z1rlawUsfxWQH1SevD6SkEa34bWtqazsvKSZ+xg68gSQojhwpD9Zx1skdoWqx11tk37g70b3bdvXwizFZ2vzmpvRbO9XvUudMWKFeGddKvoz3Z07gyLq56wNdyIDRs29HiXDBb+5Zdfwpkw6TyzZ8/Orly5UoTqyRJCiOHCkHXCOI18EtHrOHXqVJGitaxbty47fPhweDfJR0w4TftZDe/IcW4cvPP0mOMjHc9xxO9DeS/OPeBDL5Nlv40mHWHejabke/LVZPf7bX9wL8Xo0aOz27dvF6En8K68zEELIYRojrY6YX4fCj/88EM4G2VfJBtvvPFGdunSpV4f6+Cg6jipmBdffDGsoL0snKQ5T7BV75EjR4Izfuutt0IYli9fnn333XdhAoDT8h+HeXCupN25c2cR0wXP7d27N9zj+Oqrr0L8xYsXg3x+v3v16tUQVwW/p2X1mDrKfmu7aNGiUPZY55SHVSr6NtBT/HU5daGt/BfnNsF47rnnwpmPvlKrfX4TzC4D1JUlhBDDhX5/HR0P1v2FlSQrs23btnWvBhlw+ScRDMw4jRT2pfWHH37YXRZWkMgZNWpUuFeXtWvXhq1lHC55I4uy4HxWr15dpOoqK196b9q0qddPecBWxY22nqdNmxbOlj7m/v37wfkbly9fDufnn38+27p1a5hgpFarrQAHic5ff/31UAd0sXTp0qCb+Mt2dP7CCy+ENMDkhDSLFy8OceiP89tvvx3k8XU5vP/++2Hywmre2ow80fOnn34a0tSVJYQQw4bOCvhIhiRlh2HpPMRxLyYVR5ivYrmXD8LhnzXEXyun5POhk31dy8HzdfPysvhiOJaVD/jJj4CI477/mpu8rbyU3b5Otrztg6c4DiyeOO5RJ/I27DoV52W1AuoQ64Fr/zW1fSlO/gbPUW/ieS7VhoD+KL/J59rrua4sIYQYDlQ6YdET+5kNzjIGR4VD5TAHjdPAyfBMXScMlp7DvpjGCZlsc8KxfCGEEEOTyv8dLbrg3eitW7fC9jTvNgfq4y8hhBD/LYbs19H/JnyMxX/QwgHzcZQQQgjRCrQSFkIIIdqEVsJCCCFEm5ATFkIIIdqEnLAQQgjRJuSEhRBCiLaQZf8Hgj46ZhtUXy0AAAAASUVORK5CYII=\"\u003e\u003c/p\u003e\n\u003cp\u003eUnder the model assumptions, a little over half of all marketable molecules are likely to survive clinical development and make it to the marketplace.\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eThe PIPEDREAM model estimates both \u003cem\u003eFP\u003c/em\u003e and \u003cem\u003eFN\u003c/em\u003e at each phase of the pharmaceutical development process, providing a picture of decision errors during drug development. The model can be adapted to different therapeutic areas or development pipelines by adjusting observed \u003cem\u003epA\u003c/em\u003e and \u003cem\u003epM\u003c/em\u003e rates.\u003c/p\u003e\n\u003cp\u003eIn the worked example, drawing upon aggregated pharmaceutical industry data from the Pharmaceutical Benchmarking Forum\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e, the false positive rates are high, and the false negatives low, throughout the early preclinical phases. Interestingly, false positive rates are lowest and false negative rates highest during Phase 2 clinical development. This may reflect an understandable reluctance on the part of decision makers to enter the most expensive phase of clinical development with equivocal Phase 2 studies.\u003c/p\u003e\n\u003cp\u003eIn contrast, false positive rates during Phase 3 are higher. This is consistent with strong progression-seeking behaviour in Phase 3 with pharmaceutical companies preferring to risk rejection and proceed to submission and product launch, rather than terminate at so late a stage in development.\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eThe PIPEDREAM model allows us to:\u003c/p\u003e\n\u003cp\u003e\u0026bull; Identify Bottlenecks: Highlight phases with high \u003cem\u003eFN\u003c/em\u003e rates where potentially marketable molecules are being lost.\u003c/p\u003e\n\u003cp\u003e\u0026bull; Optimize Decision-Making: Provide insights into how changes in progression criteria could reduce \u003cem\u003eFP\u003c/em\u003e and \u003cem\u003eFN\u003c/em\u003e rates.\u003c/p\u003e\n\u003cp\u003e\u0026bull; Benchmark Performance: Compare the efficiency of different development pipelines or therapeutic areas.\u003c/p\u003e\n\u003cp\u003eOf course, there are limitations both to the data and to the PIPEDREAM model itself. The estimated \u003cem\u003eFP\u003c/em\u003e and \u003cem\u003eFN\u003c/em\u003e rates are based upon historical data derived from pharmaceutical companies participating in the Pharmaceutical Benchmarking Forum aggregated across all therapeutic areas. More recent data on progression and marketing rates across all phases of discovery and development may not be available, but the model is flexible and permits estimation of decision risks for the clinical phases of the development process for which more recent data are available for individual therapeutic areas.\u003c/p\u003e\n\u003cp\u003eThe PIPEDREAM model largely sidesteps the issue of independence of the phases by modelling the probabilities of market launch at each phase \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003econditional\u003c/span\u003e upon progression to that phase. However, the PIPEDREAM model incorporates potential dependencies by including an autoregressive correlation structure with low, moderate, and high dependencies between phases.\u003c/p\u003e\n\u003cp\u003eThe choice of a stationary correlation structure where \u003cem\u003ethe pattern of dependencies between phases remains constant\u003c/em\u003e throughout the entire development pipeline. The AR1 model assumes that the correlation between Phase 1 \u0026amp; 2 \u003cem\u003eFP\u003c/em\u003e rates and the correlation between Phase 6 \u0026amp; 7 \u003cem\u003eFP\u003c/em\u003e rates are both equal with \u003cem\u003e\u0026rho;^1\u0026thinsp;=\u0026thinsp;\u0026rho;\u003c/em\u003e, despite being different phase pairs.\u003c/p\u003e\n\u003cp\u003eThis does not rule out carryover between phases. For example, we might expect that a drug that sails through, say, Phase 1 is more likely to progress through to Phase 2, than a drug which narrowly makes it through Phase 1. In this case the model would underestimate progression rates in Phase 2. Conversely, the Winner\u0026rsquo;s Curse might mean that a drug that sails through Phase 2 with an inflated overestimate of effect size is less likely to make it through Phase 3 as the model will overestimate Phase 3 progression rates.\u003c/p\u003e\n\u003cp\u003eIn addition, real-world development processes may exhibit phase-dependent decay with later phases (e.g., P3 to STL) showing stronger correlation persistence than early phases (TTH to HTL). In addition, there may be regime changes at key points. Major phase transitions (e.g., preclinical to clinical) could alter dependence patterns. In addition, external influences such as changes in trial designs over the last ten years or shifting regulatory landscapes might modify correlations over time.\u003c/p\u003e\n\u003cp\u003eWhile a stationary correlation structure enables tractable modelling, it may oversimplify complex phase transition dynamics, underestimate uncertainty in later phases, and might best be improved with phase-specific correlation data.\u003c/p\u003e\n\u003cp\u003eFurthermore, the model assumes that the final, STL, phase is an absorptive state with known \u003cem\u003eFP\u003c/em\u003e and \u003cem\u003eFN\u003c/em\u003e rates where \u003cem\u003eFP\u003c/em\u003e\u003csub\u003eSTL\u003c/sub\u003e = 1 - \u003cem\u003epM\u003c/em\u003e\u003csub\u003eSTL\u003c/sub\u003e and \u003cem\u003eFN\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0. However, this is not critical. The model allows users to experiment with other choices for \u003cem\u003eFP\u003c/em\u003e\u003csub\u003eSTL\u003c/sub\u003e and \u003cem\u003eFN\u003c/em\u003e\u003csub\u003eSTL\u003c/sub\u003e.\u003c/p\u003e\n\u003cp\u003eNevertheless, this analysis gives some insight into recent drug development strategies. In particular, the pharmaceutical development pipeline may be richer in marketable drugs than currently supposed. During the period through to 2010, just 4% of projects entering development survived the pharmaceutical development process and made it to market. Many of these marketable drugs may be lost during Phase 2 trials. This analysis supports the strategic targeting of Phase 2 trials and the need to better power Phase 2 studies to reduce false negatives.\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eAt least for the clinical phases, more recent data\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e suggest that \u003cem\u003epA\u003c/em\u003e and \u003cem\u003epM\u003c/em\u003e have changed little since the Pharmaceutical Benchmarking Forum of 2010. We welcome more recent data for all the phases of the current pharmaceutical research and development process.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eKola, I. \u0026amp; Landis, J. Can the pharmaceutical industry reduce attrition rates? Nat Rev Drug Discov 3, 711-715, doi:10.1038/nrd1470 (2004)\u003c/li\u003e\n\u003cli\u003ePaul, S. M. et al. How to improve R\u0026amp;D productivity: the pharmaceutical industry\u0026apos;s grand challenge. Nat Rev Drug Discov 9, 203-214, doi:10.1038/nrd3078 (2010). \u003c/li\u003e\n\u003cli\u003eLindborg, S. R., Persinger, C. C., Sashegyi, A., Mallinckrodt, C. \u0026amp; Ruberg, S. J. Statistical refocusing in the design of Phase II trials offers promise of increased R\u0026amp;D productivity. Nat Rev Drug Discov 13, 638-640, doi:10.1038/nrd3681-c1 (2014). \u003c/li\u003e\n\u003cli\u003eLendrem, D. W. et al. Progression-seeking bias and rational optimism in research and development. Nat Rev Drug Discov 14, 219-221, doi:10.1038/nrd4320-c1 (2015). \u003c/li\u003e\n\u003cli\u003eDimitri, N. An assessment of R\u0026amp;D productivity in the pharmaceutical industry. Trends in pharmacological sciences 32, 683-685, doi:10.1016/j.tips. 2011.09.005 (2011). \u003c/li\u003e\n\u003cli\u003eOwens, P. K. et al. A decade of innovation in pharmaceutical R\u0026amp;D: the Chorus model. Nat Rev Drug Discov 14, 17-28 doi:10.1038/nrd4497 (2015). \u003c/li\u003e\n\u003cli\u003eLendrem, D. W. \u0026amp; Lendrem, B. C. Torching the Haystack: modelling fast-fail strategies in drug development. Drug Discov. Today 18, 331-336 (2013). \u003c/li\u003e\n\u003cli\u003ePeck, R. W. Driving earlier clinical attrition: if you want to find the needle, burn down the haystack. Considerations for biomarker development. Drug Discov. Today 12, 289-294 (2007). \u003c/li\u003e\n\u003cli\u003ePeck, R.W. et al. Why is it hard to terminate failing projects in pharmaceutical R\u0026amp;D? Nat Rev Drug Discov 14, 1-2, doi10.1038/nrd4725 (2015)\u003c/li\u003e\n\u003cli\u003eScrucca L \u0026ldquo;GA: A Package for Genetic Algorithms in R.\u0026rdquo; Journal of Statistical Software, 53(4), 1-37. doi:10.18637/jss.v053.i04 https://doi.org/10.18637/jss.v053.i04 (2013).\u003c/li\u003e\n\u003cli\u003eScrucca L \u0026ldquo;On some extensions to GA package: hybrid optimisation, parallelisation and islands evolution.\u0026rdquo; The R Journal, 9(1), 187-206. doi:10.32614/RJ-2017-008 https://doi.org/10.32614/RJ-2017-008 (2017).\u003c/li\u003e\n\u003cli\u003eMarius Hofert, Ivan Kojadinovic, Martin Maechler and Jun Yan \u0026ldquo;copula: Multivariate Dependence with Copulas. R package version 1.1-5\u0026rdquo; https://CRAN.R-project.org/package=copula (2025). \u003c/li\u003e\n\u003cli\u003eJun Yan Enjoy the Joy of Copulas: With a Package copula. Journal of Statistical Software, 21(4), 1-21. https://www.jstatsoft.org/v21/i04/ (2007).\u003c/li\u003e\n\u003cli\u003eMarius Hofert, Martin Maechler Nested Archimedean Copulas Meet R: The nacopula Package. Journal of Statistical Software, 39(9), 1-20. URL https://www.jstatsoft.org/v39/i09/ (2011)\u003c/li\u003e\n\u003cli\u003eCorporation M, Weston S doParallel: Foreach Parallel Adaptor for the \u0026apos;parallel\u0026apos; Package_. R package version 1.0.17, https://CRAN.R-project.org/package=doParallel (2022).\u003c/li\u003e\n\u003cli\u003eWickham H, Fran\u0026ccedil;ois R, Henry L, M\u0026uuml;ller K, Vaughan D dplyr: A Grammar of Data Manipulation_. R package version 1.1.4, https://CRAN.R-project.org/package=dplyr (2023).\u003c/li\u003e\n\u003cli\u003eR Core Team R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/ R version 4.4.2 (2024-10-31 ucrt)\u003c/li\u003e\n\u003cli\u003eKnowledge Portal on Innovation and Access to Medicines, Research Synthesis: Time and Success Rates of Pharmaceutical R\u0026amp;D. https://www.knowledgeportalia.org/r-d-time-and-success-rate (2020)\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[{"identity":"98a7663e-897a-48cc-9900-11fbf9624307","identifier":"10.13039/501100000774","name":"Newcastle University","awardNumber":"NA","order_by":0}],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Newcastle University","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Strategy, Drug Development, Costs, Risks, False Positives, False Negatives, False Discoveries, Opportunity Costs, Decision Making, PIPEDREAM","lastPublishedDoi":"10.21203/rs.3.rs-6198126/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6198126/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe PIPEDREAM model estimates false positive and false negative rates for each phase of the drug development process using data on attrition and marketing rates for each phase of development. We illustrate the model using aggregated historical data for the pharmaceutical industry from 1997-2007. These data suggest that false positive rates were high, and false negative rates low, throughout the early phases of development during this period. False positive rates were lowest and false negative rates peaked during Phase 2 clinical development. Furthermore, there was evidence of progression-seeking behaviour in Phase 3, with more molecules progressing to submission than might be expected. \u003cbr\u003e\n\u003c/p\u003e","manuscriptTitle":"PIPEDREAM: Pharmaceutical Innovation, Progression Errors, Development Risks, Efficiency and Attrition in Medicine","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-29 14:36:54","doi":"10.21203/rs.3.rs-6198126/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"9f229b92-9b93-4325-9a1d-2b254ca32320","owner":[],"postedDate":"July 29th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":52273721,"name":"Translational Medicine"}],"tags":[],"updatedAt":"2025-07-29T14:36:54+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-29 14:36:54","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6198126","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6198126","identity":"rs-6198126","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}