Costs of Pharmaceutical R&D
v1.0 researched and written by Marcela Vieira, edited by Suerie Moon, last updated January 2020
The literature on costs for product research and development (R&D) in the pharmaceutical sector is considerable*. The vast majority of the literature focuses on development costs (clinical stage), with less information available on research costs (discovery and pre-clinical stage). Almost all studies are related to new drugs, with scarce data available for other types of medical products and devices. And almost all data are from pharmaceutical companies, with only four papers providing data from other types of organizations conducting drug development.
Cost and research and development and pharmaceutical.
Search was conducted using a combination of search mechanisms, mainly in English, with no specific time period of publication.
Contribute to improve and update this page!
We particularly welcome suggestions on gaps in the reviews and on interesting new research.
Synthesis of the literature
The product development process in the health and pharmaceutical sector is characterized as complex, lengthy and costly. R&D costs influence decisions and policy options about how to best incentivize innovation to meet health needs and how to make end products available at affordable prices. This research synthesis focuses on the cost aspect of the process. R&D costs estimations for the development of new drugs range widely, from $43.4 million to $4.2 billion. Estimations are usually from preclinical stage to market approval and most do not include post approval costs, or costs for the discovery stage, although estimates for these are available in the literature. The majority of the studies provide figures for out-of-pocket costs and for capitalized costs (cost of capital/opportunity costs/time costs), which constitutes from 14-51% of total costs depending on the capitalization rate, timeframes and methodology adopted. Table 1 provides a summary of the studies.
There is a variety of methods and data sources (many confidential) used in the studies, which can explain the wide range of cost estimates. The type of expenses that are included as R&D costs can also vary and are not always explicitly mentioned in the calculations (such as costs of failures, expenses related to marketing of the product or tax deductions). A deep analysis of the studies is beyond the scope of this research synthesis, which aims to provide an overview of the topic and highlight key findings from the range of studies, including variables that can affect R&D costs and main drivers.
Most of the studies focus on the development of new drugs (“new molecular entities” or “new chemical entities”) by large pharmaceutical companies, and exclude cost to develop variations of existing drugs (such as combinations or reformulations). Some provide information on product development by non-commercial actors, but there is a need for further studies on this. There is also a need for more information about costs related to the development of other health technologies beyond drugs.
Table 1. Summary of recent estimates of new drug development costs (in millions of USD)
Summary of the contents
This research synthesis is organized by the following topics:
- New drugs – overall
- New drugs by therapeutic area
- New drugs by company size
- Non-commercial product developers
- Other types of health products
1) Estimations of R&D costs
New drugs – overall
We identified two studies that reviewed the literature on R&D costs for drug development, one published in 2011 and the other in 2012, and one focusing on the costs of clinical trials published in 2018.
Mestre-Ferrandiz et al. (2012), in a study published by the United Kingdom Office of Health Economics (OHE), identified 10 studies with estimates of R&D costs for new drugs conducted by industry, concluding that the available literature shows “an increase in [average, capitalized] costs from £125 million (USD 199 million) per new medicine in the 1970s to £1.2 billion (USD 1.9 billion) in the 2000s”. Figures are in 2011 dollars.
Morgan et al. (2011) reviewed 13 articles containing original estimates of the cost of drug development published from 1980 to 2009. They found estimates ranging from USD$92 million cash (USD$161 million capitalized) to USD$883.6 million cash (USD$1.8 billion capitalized), expressed in 2009 dollars. The authors suggested that differences in methods, data sources, and time periods explain some of the variation in estimates. They highlight that lack of transparency limits many studies, especially as “confidential information provided by unnamed companies about unspecified products forms all or part of the data underlying 10 of the 13 studies” and concluded that “despite three decades of research in this area, no published estimate of the cost of developing a drug can be considered a gold standard”.
Speich et al. (2018a) conducted a systematic review on costs of randomized clinical trials (RCTs), the most expensive stage of drug development. They found 56 articles, none of which provided empirical cost data for all aspects of a trial, but provided information on several aspects of the trial, such as cost data of different development phases, recruitment costs or site-specific costs, or aggregated overall costs. Results show that “the median costs per recruited patient were USD 409 (range: USD 41–6,990)” and overall costs ranged from USD 0.2–611.5 million per RCT. All cost data are in 2017 dollars. The authors highlight that the studies use different methodologies, and that 75% of the articles did not provide detailed information on the methodology and underlying data used in the calculations.
Deloitte Center for Health Solutions have published over the past 10 years annual reports on pharmaceutical R&D. The last report (Deloitte, 2019), estimates the average R&D costs to develop a new drug at USD 1,981 million for the original cohort and USD 2,422 million for the extension cohort, inclusive of costs of failures. The original cohort is composed of 12 largest biopharmaceutical companies (ranked by R&D expenditures in 2009) and the extension cohort is comprised of 4 “mid-to-large, more specialized companies” (not specified). The 2019 report also provides an overview of the estimations from the previous reports summarized in the figure below. The estimation is calculated based on total R&D spending by the cohort companies, which in 2019 amounted to USD 79 billion, and the total of assets in late-stage pipeline (defined as “assets that are filed, in Phase III or Phase II with breakthrough therapy designation as of 30th April each year”), which included, in 2019, 183 products for the original cohort and 30 for the extension cohort. In 2019, that included 43% of small molecule drugs (down from 67% in 2010), 37% of antibody therapies (up from 15% in 2010) and 20% of other modalities, including cell and gene therapies, protein-based therapies, vaccines and synthetic peptides (proportion stable over the years). The costs are collected from “publicly available information from audited annual reports or readily available from third-party data providers” and includes costs from discovery to launch. The methodology used to calculate average R&D cost per product is not specified in detail, and it is unclear whether total costs are out-of-pocket or capitalized, or if they have been adjusted for inflation.
Source: Deloitte Center for Health Solutions, 2019, p. 13.
DiMasi and co-authors at the Tufts Center conducted a series of studies to estimate private sector R&D costs of drugs. The latest study published in 2016 (DiMasi et al., 2016) estimated the average (mean) pre-tax out-of-pocket cost to be USD$1,395 million (2013 dollars) and USD$2,558 million (2013 dollars) after capitalization at 10.5%. Time costs (or opportunity costs) accounts for 45% of total costs. Breaking down the total costs between preclinical and clinical stages, the estimated preclinical out-of-pocket costs are $430 million and $1,098 million capitalized and for clinical stage out-of-pocket costs are $965 million and $1,460 million capitalized. Therefore, preclinical costs were estimated to represent 32% of total out-of-pocket costs and 42% of total capitalized costs. The study also provides an estimate of post-approval R&D costs at USD$312 million after capitalization, which would increase the cost estimate to USD$2,870 million (2013 dollars) if included. The calculation includes costs related to compounds abandoned during the development process (cost of failures). The calculation does not consider public subsidies or tax deductions/credits tied to R&D expenditures, which would reduce net R&D costs for a company. The estimations were based on a confidential survey through which 10 multinational pharmaceutical companies (not named) of varying sizes self-reported information on R&D costs of 106 new drugs, out of which 19 were biologics (the drugs are not specified). Full R&D costs were not provided for all of the 106 drugs, and the sample size per phase included 78 drugs for preclinical, 97 for phase I, 78 for phase II and 42 for phase III. The sample included self-originated compounds with first test in human anywhere in the world from 1995-2007 and development costs occurred through 2013, including preclinical stage from the point of compound syntheses/isolation until market approval (therefore not including pre-synthesis discovery costs). It is noted that the clinical costs estimations include not only clinical trial costs per se, but also “R&D costs not directly related to the trials that were incurred during the phase, such as those for long-term animal testing, chemistry, manufacturing and controls (CMC), and company R&D overhead”. It is also noted that total cost estimations are inclusive of “non-molecule related costs required to run an R&D organization”.
Using similar methodology, in the 2003 study, DiMasi et al. (2003) investigated the costs related to 68 new drugs using self-reported data from 10 multinational pharmaceutical firms. The sample drugs entered clinical development stage anywhere in the world between 1983 and 1994 and development costs were included until 2000. The authors estimated the mean pre-tax pre-approval out-of-pocket cost per new drug in USD$ 403 million (2000 dollars) or USD$ 802 million (2000 dollars), after capitalization at 11%, including the costs of failures. In the 1991 study, the same group of authors (DiMasi etal., 1991) estimated the R&D costs of 93 new chemical entities (including costs of failures) using self-reported data from a survey of 12 US-owned pharmaceutical companies. The selected drugs initiated clinical trials between 1970 and 1982. The estimated pre-tax out-of-pocket cost was USD$114 million (1987 dollars) and US$231 million (1987 dollars) when capitalized at 9%.
Light and Warburton (2011) analyzed the R&D costs estimated by DiMasi et al. (2003) and argued that the costs were inflated to support industry-claims of high risks and costs involved in the development process used to justify high prices of medicines. They formulate several critiques to the methodology used in the DiMasi study, such as inflated trial costs, exaggerated time for each development phase and capitalization of costs at a high rate. The authors adjust the data to account for those factors and suggest that actual R&D costs might be much lower. There are several estimations adjusted for different factors, with the lowest being median net out-of-pocket costs of USD$43.4 million per new drug (including 50% tax reductions), 18 times lower than the figure in the analyzed study (USD$802 million).
Mestre-Ferrandiz et al. (2012), beyond reviewing the available literature on R&D costs, also presented a new estimative at USD 1,506 million (in 2011 dollars, capitalization at 11%). Out-of-pocket costs were calculated at USD 899 million (2011 dollars). The calculation was “based on previously unpublished information collected by CMRI in confidential surveys”. Costs were estimated based on reported R&D expenditures of the companies included in the surveys and the 97 projects in development from 1997-2002.
PricewaterhouseCoopers (2012), calculated the average R&D costs of new drugs to be USD 4.2 billion per molecule in the period 2007-2011 and USD 2.8 billion from 2002-2006. The calculation was based on the total number of new drugs (including biologics) approved by the FDA from 2002-2011, which amounts to 308, and how much the pharmaceutical industry invested in R&D in the same period, totaling USD 1.1 trillion. Estimates seems to be out-of-pocket costs, not capitalized and not adjusted to inflation, but the methodology is not clear in the study. They also reported average industry R&D expenditures to be distributed as following: 7.1% for target selection/validation; 21.5% for screening/lead optimization; 9.2% for proof of mechanism/phase I; 17.4% for proof of concept/phase II; 39.8% for phase III and 5% for approval.
Adams and Brantner (2010) calculated expenditures related to the clinical stage of development of new drugs using publicly available data from 183 publicly traded companies in the pharmaceutical industry in the US in a 12-year period (1989–2001). They estimated that average costs for the clinical stage was USD$27 million per year, being USD$17 million for Phase I trials, USD$34 million for Phase II and USD$27 million for Phase III, which multiplied by average phase durations results in estimates of $24 million, $86 million, and $61 million for Phases I, II, and III, respectively. The authors concluded that their estimated costs were higher than suggested in previous studies. The same authors also conducted a similar study in 2006 (Adams and Brantner, 2006), estimating total clinical stage costs at a range of USD$500 million to USD$ 2 billion per drug, depending on the therapeutic class and on the company. Figures from both studies are in 2000 dollars and capitalized at 11%.
Paul et al. (2010) estimated R&D costs of new drug development, from preclinical (target-to-hit) to launch, based on assumptions of success rates, timeframes and costs per phase based on industry benchmarks and cost data from Eli Lilly and Company. The total out-of-pocket cost to launch one NME is estimated at USD 873 million, and capitalized cost at USD 1,778 million (at 11%). See Table 1 for costs per phase. The authors note that the estimate does not include discovery research, post-launch expenses or overheads.
Public Citizen (2001) conducted an analysis drawing on publicly available US Securities and Exchange Commission (SEC) R&D filings for all major pharmaceutical firms and calculated R&D costs, after-tax deductions, to be from USD 57 million to USD 71 million for the average new drug brought to market in the 1990s, including failures. Pre-tax estimations were calculated to be an average of USD 163 million per new drug approved. The calculations are based on PhRMA data for R&D spending (USD 139.8 billon in the 1990s) and the number of NMEs approved by the US FDA in the period (857). Figures are in 2000 dollars.
The US Congress Office of Technology Assessment – OTA (1993) calculated that “the average aftertax R&D cash outlay for each new drug that reached the market in the 1980s was about $65 million (in 1990 dollars)”. And “the full aftertax cost of these outlays, compounded to their value on the day of market approval, was roughly $194 million (1990 dollars)”. The study estimated that tax savings and credits reduced out-of-pocket costs by nearly 50 per cent.
The literature review conducted by Mestre-Ferrandiz et al. (2012) also included information on two studies that investigated R&D cost in the period from 1960s to 1980s and updated their estimations to 2011 prices. According to the review, Wiggins, in a study published in 1987, investigated the period from 1970-1985 and calculated the development costs of a new drug to be of USD 226m (fully capitalized) (in 2011 prices). The calculation was based on the total number of new molecular entities (NMEs) approved by FDA in the period and an estimation of R&D spending on new drug development in previous years at industry level. Hansen, in a study published in 1979, analyzed project-specific data on 67 products that entered clinical testing between 1963 and 1975 and were approved for marketing starting around 1970. Hansen estimated the cost to be US$199m (in 2011 dollars).
Focusing on the preclinical stage, Horvath (2010) estimated the costs of development for small molecule drugs and biologics. The estimations were based on a simulation of preclinical studies costs conducted by the author in 2004. Preclinical studies necessary to obtain the IND – Investigational New Drug Application for small molecule drugs were estimated to last 10 months and cost USD 1.2 million, including manufacturing costs estimated at USD 40,000. For biologics, they were estimated to last 17 months and cost USD 2.6 million, including USD 1.6 million cost to manufacture the material. The author also provides calculations for preclinical studies necessary to support each phase of clinical development up to registration. Total supporting preclinical studies were estimated to require 73 months and cost USD 7 million for drugs and 47 months and USD 6.3 million for biologics.
New drugs by therapeutic area
The literature shows that the costs to develop a new medicine can vary significantly according to the therapeutic area. The above-mentioned literature review conducted by Mestre-Ferrandiz et al. (2012) looked into costs by therapeutic area and concluded that the most expensive are neurology, respiratory and oncology, and the less expensive are anti-parasitics and drugs to treat HIV/AIDS due to differences in success rates and development times.
Sertkaya et al. (2014), in a report prepared by Eastern Research Group, Inc. for the US Department of Health and Human Services, used “aggregate data from three proprietary databases on clinical trial costs provided by Medidata Solutions”, to estimate the costs of each phase of the clinical stage of drug development for 13 different therapeutic area. Costs include only information from industry-sponsored trials conducted in the US. A summary of the estimations is available in the figure below. The authors conclude that “the therapeutic area with the highest average per-study costs across Phases I, II and III is pain and anesthesia ($71.3 million) followed by ophthalmology ($49.9 million) and anti-infective ($41.3 million) trials. Conversely, trials in dermatology, endocrinology, and gastroenterology have the lowest overall costs across the same three phases”. They also looked at costs for Phase IV studies, usually not included in R&D costs estimations, and concluded that “average Phase IV study costs are equivalent to those of Phase III costs but are much more variable across different therapeutic areas than Phase III costs”. Costs are capitalized at 15%. Beyond per-study costs, Sertkaya et al. (2014) also provide estimations of costs for several components grouped into per-study costs, per-patient costs, and per-site costs, adding additional costs for site overhead (estimated at 25% of per-study costs) and all other additional costs not captured by those categories (estimated at additional 30% of all the costs combined).
Source: Sertkaya et al., 2014, p. 3-3.
The study by Adams and Brantner (2010) presents estimates of expenditures on drug development according to their therapeutic categories. Their findings show that cardiovascular, dermatological, genitourinary, anticancer and neurological drugs are all above average and that biotech drugs are below average. They also highlighted that new formulations of existing drugs have substantially smaller expenditure than the average. In their 2006 study (Adams and Brantner, 2006), the same authors suggested that variation in R&D costs by therapeutic class might be due to differences in success rates and length of trials, and not on actual spending.
Prasad and Mailankody (2017) examined the R&D costs to bring 10 cancer drugs to market and revenues after approval, using data from 10 companies’ filings to the US Securities and Exchange Commission. The selection criteria were drugs developed by companies that had no other product in the market, therefore, mostly small companies. The estimation included all R&D costs reported by the companies, therefore inclusive of costs of other compounds that did not reach regulatory approval (cost of failures). The authors found that the median time to develop a drug was 7.3 years (range, 5.8-15.2 years) and estimated the median out-of-pocket cost of development at USD$648.0 million (range, USD$157.3 million to USD$1,950.8 million). If capitalized for a 7% opportunity costs, median cost was USD$757.4 million (range, USD$203.6 million to USD$2,601.7 million). The authors concluded that the estimated cost to develop a cancer drug of USD$648.0 million is significantly lower than prior estimates (from USD$320.0 million to USD$2.7 billion). Figures are in 2017 dollars.
Jayasundara et al. (2019) investigated the differences in clinical trial costs to develop orphan drugs and non-orphan drugs. They randomly selected 100 new orphan drugs and 100 non-orphan drugs approved by the US FDA between 2000 and 2015 (new indications, new formulations, new manufactures and new dosage forms of already marketed drugs were excluded). They used the clinicaltrials.org database to identify all clinical trials, numbers of patient enrolled and trial duration for each development phase (phase I to III), amounting to a total of 1,163 trials (602 for orphan drugs and 561 for non-orphan drugs). The costs were calculated based on per-trial cost averages reported at PhRMA 2015 for non-orphan drugs and at EvaluatePharma 2015 for orphan drugs, and adjusted to probability of success based on DiMasi 2016, to include the cost of failures. They found that “the out-of-pocket clinical costs per approved orphan drug to be $166 million and $291 million (2013 USD) per non-orphan drug. The capitalized clinical costs per approved orphan drug and non-orphan drug were estimated to be $291 million and $412 million respectively”. They also conducted a separate analysis only for new molecular entities (NMEs), which represents 54 out of the 100 non-orphan drugs and 74 of the 100 orphan drugs. This resulted in an “estimated capitalized cost per approved non-orphan NME was $489 million and the capitalized cost per approved orphan NME was $242 million”, meaning that the costs to develop orphan drugs are about 50% lower than non-orphan drugs, despite high prices of orphan drugs being attributed to the high cost of drug development. Figures are in 2013 dollars and capitalization was at 10.5% rate.
The above-mentioned paper by Dimasi et al. (2016) provided cost information per clinical phase of development broken-down between small molecules and large molecules (biologics). Out of the 106 drugs included in the study, 19 were large molecules and 87 small molecules. Mean costs for phase I, were USD 23.93 million for large molecules and USD 25.53 million for small molecules; for phase II, USD 91.86 million for large molecules and USD 50.40 million for small molecules; for phase III, USD 281.13 million for large molecules and USD 245.80 for small molecules. After adjusting for approval success rates, which were much higher for large molecules (28.9% compared to 9.3%), clinical costs for the development of small molecules were 44.5% higher than for large molecules.
DiMasi et al. (2016) also found that average development costs are lower for priority drugs, as rated by the US FDA, than for standard drugs. Out-of-pocket costs for clinical stage per approved compound were USD 554 million for standard-rated and USD 385 for priority-rated compounds, while capitalized costs were USD 782 million and USD 489 million respectively. The authors suggests that while priority-rated compounds would tend to be costlier for breaking more new scientific ground, standard-rate compounds usually requires large trials to demonstrate superiority or non-inferiority to similar compounds already in the market. The authors conclude that “our results can be viewed as supportive, but not conclusive, evidence of higher costs for drug with lower therapeutic significance ratings”.
DiMasi and Grabowski (2007) estimated the costs of developing biologic drugs by biotech companies, focusing specifically on recombinant proteins and monoclonal antibodies [mAbs], and compared to the costs of developing small-molecule drugs by traditional pharmaceutical companies. They found average out-of-pocket cost estimates per approved biopharmaceutical to be of USD 559 million, separated in USD 198 million for the preclinical stage and USD 361 million for the clinical stage. After capitalization, the average costs amount to a total of USD 1,241 million, being USD 615 million for preclinical and USD 626 for clinical development. They concluded that total out-of-pocket cost per approved biopharmaceutical was lower than for the traditional pharmaceutical companies (USD 559 million vs USD 672 million) and that capitalized cost was nearly the same (USD$1,241 million versus USD$1,318 million). After adjusting for approval success rates (higher for large molecules), the authors concluded that clinical costs per approved compound were 44.6% higher for small molecules than for large molecules, while total development costs were similar for the two types of drugs. Figures are in 2007 dollars.
DiMasi et al. (2004) conducted a study breaking down R&D costs data for the clinical stage of development from 68 drugs (inclusive of failures) into four therapeutic classes. Data was provided by 10 large pharmaceutical companies. All figures are expressed in 2000 US dollars. Overall average of out-of-pocket cost was estimated to be $282 million. The costs related to anti-infective drugs were considerably above average ($362 million) and analgesic/anesthetic drugs were modestly below average ($252 million), while for cardiovascular ($277 million) and CNS -central nervous system ($273 million) the costs were close to the overall average. Considering capitalized costs, the overall average was $466 million, costs were slightly lower for CNS ($464 million) and for cardiovascular ($460 million) drugs, significantly lower for analgesic/anesthetic drugs ($375 million) and higher for anti-infective drugs ($492 million). The same study also estimated worldwide sales for all new drugs approved in the United States from 1990 to 1994 over a 20-year period, using data from IMS Health. The mean net values of sales were $2,434 million for all drugs, $1,080 million for analgesic/anesthetic drugs, $2,199 million for anti-infective drugs, $3,668 million for cardiovascular drugs, and $4,177 million for CNS drugs. The authors concluded that the sales did not correlate well with average development costs, but “the results are still consistent with a model of firm behavior that posits that R&D efforts will generally shift toward high net return, and away from low net return, therapeutic areas”.
In a similar study published in 1995, the same group of authors (DiMasi et al., 1995a) found that the overall mean capitalized costs per approved NCE including the cost of failures were USD$93 million, being USD$70 million for anti-infective drugs, USD$98 million for cardiovascular, USD$103 million for neuropharmacological and USD$163 million for anti-inflammatory drugs. If costs of unsuccessful projects were excluded, then the mean costs ranged from USD$7.1 million (for topical steroids) to USD$66.7 million (for cardiovascular) (all in 1993 US dollars). The authors highlighted that “phase attrition and approval rates are the most important sources of variability in total clinical period costs between therapeutic categories”. In relation to sales, the authors concluded that “development cost estimates by therapeutic category did not correlate strongly with US sales in the fifth year of marketing”, mentioning that cardiovascular drugs had revenues much higher than average, while having on-average development costs, while nonsteroidal anti-inflammatory drugs had average revenues, but much higher than average development costs.
New drugs by company size
The size of the company conducting the development process is also appointed as a factor that can affect the costs of R&D. The above-mentioned report by Deloitte Center for Health Solutions (2014), concluded that “the larger the company, by revenue or R&D spend, the greater the cost to develop each asset and the lower the returns.” The literature review conducted by Mestre-Ferrandiz et al. (2012) concluded that “results of research on the impact of firm size on R&D productivity and R&D costs are mixed. The evidence from the 1990s and early- to mid-2000s seems to suggest that size matters: multiple tangible and intangible assets are associated with fully integrated organisations, where core capacities can be important across diseases. It remains unclear, however, whether R&D productivity is greater for smaller companies than for traditional “big pharma””.
Adams and Brantner (2006) also explored the relationship between company size and R&D costs. The authors note that “it has been argued that larger companies have economies of scale and scope in drug development that might be associated with lower development costs”. They explained that this might be related to the method used to classify the company “when an ex post measure of size (Top 10 by 2001 income) is used, the average drug from a large firm has a cost much lower than the overall average. However, when ex ante measures of size are used, the cost of the average drug from a large firm is larger than the cost for the overall average drug”. Their findings showed that “drugs from firms that had the largest number of drugs in development had an average capitalized cost of $992 million—some $124 million more than the average drug”. Therefore, they conclude that the “results do not support the claim that larger firms tend to produce lower-cost drugs”, contrasting with previous work that found that drugs from small firms tend to have higher costs than drugs from larger firms.
DiMasi et al. (1995b) analyzed the relationship between R&D costs and the company size, using self-reported data from 12 US pharmaceutical companies. The companies were grouped into three groups according to their sales level. The authors concluded that “the R&D cost per new drug approved in the US is shown to decrease with firm size, while sales per new drug approved are shown to increase markedly with firm size”.
Non-commercial product developers
DNDi – Drug for Neglected Diseases Initiative (2019), a Product Development Partnership (PDP), published an updated cost estimate to develop and register new combinations or new formulations of existing treatments for EUR 4-32 million, and a new chemical entity for EUR 60-190 million. Costs are based on their own experience of drug development since their foundation in 2003, inclusive of costs of failures and “do not include post-registration studies and access costs, nor in-kind contributions from pharmaceutical partners” (it is mentioned that in-kind contributions amounted to 12.5% of DNDi total expenditures). Costs are “fully-loaded” (including management and indirect costs) out-of-pocket expenses and have not been capitalized. It is not clear if costs have been adjusted for inflation. The report also provides out-of-pocket costs for each stage of development from discovery to registration for 8 drugs: 3 “existing drugs without new formulation”, 3 “existing drugs with new formulation” and 2 “new chemical entities”. The costs range from EUR 0.1-22.6 million for discovery and preclinical; EUR 1.5-9.9 million for phase I; EUR 3.3-43.8 million for phase II, III and registration and totals from EUR 4-58 million. In 2014, DNDi had estimated its cost of development at EUR 6-20 million for an improved treatment and EUR 30-40 million for a new chemical. If the costs of failures are included, the cost range of an improved treatment would amount to EUR 10-40 million and EUR 100-150 million for a new chemical entity (DNDi, 2014).
The Global Alliance for Tuberculosis Drug Development (2001), a not-for-profit organization created to accelerate the discovery and development of new TB drugs, estimated the costs of successfully developing a new chemical entity to treat TB to be from USD$ 36.8 million to USD$ 39.9 million (excluding costs of failure). This estimation covers preclinical development ($4.9 million to $5.3 million), pharmaceutical development (at least $5.3 million), and Phases I through III of clinical development ($26.6 million). Including the costs of unsuccessful projects, the estimates of the costs of developing an NCE are from USD$ 76 million to USD$ 115 million. These estimates do not include the costs of discovery, which are estimated to range from USD$ 40 million to USD$ 125 million (including the costs of failure), leading to estimated costs of discovering and developing a new anti-TB drug (including the costs of failure) of between USD$ 115 million and USD$ 240 million.
Speich et al. (2018b) provided a retrospective assessment of costs related to two randomized clinical trials conducted in the academic setting. The “Prednisone Trial” (community-acquired pneumonia) was conducted at seven centers in Switzerland from 2009 to 2015 and involved a total of 802 patients for a total time of 59 months. The “Oxantel Trial” (intestinal worms) was conducted within a collaboration of researchers from Switzerland and Tanzania and took place on Tanzania in 2012 involving 480 children over a period of 2 months. The overall costs for the Prednisone-Trial were calculated to be USD 2.3 million, and in the Oxantel-Trial were USD 100,000. The costs stratified by categories of the Prednisone-Trial were “USD 231,347 (10.1%) for trial conception, planning, and preparation, USD 1,938,958 (84.2%) for patient enrollment, treatment, and follow-up, and USD 129,518 (5.6%) for the time after last patient out”. For the Oxantel-Trial “salary costs accounted for the largest amount of the overall costs (i.e., USD 84,447; 84.1%). Costs for the trial conception, planning, and preparation phase were USD 26,437 (26.3%), USD 45,016 (44.8%) for patient enrollment, treatment, and follow-up, and USD 28,537 (28.4%) for the after last patient out phase”. Costs seen to be out-of-pocket costs, therefore not capitalized, and not adjusted to inflation.
Other types of health products
Light et al. (2009) estimated R&D costs for two new rotavirus vaccines [RotaTeq (Merck) and Rotarix (GlaxoSmithKline or GSK)] based on publicly available information from 5 sources - the U.S. Patent and Trademark Office, the U.S. SEC EDGAR database, Medline, periodicals, and corporate websites – and complemented by interviews with senior informants. They estimated out-of-pocket cost for Phase I trials at $66,000–$264,000 for Merck and $37,600–$150,400 for GSK; Phase II trials at $0.9 million–$1.2 million for Merck and $1.8 million–$2.4 million for GSK, and Phase III trials at $136.1 to $204.1 million for Merck, and $126.4 to $189.7 million for GSK. Total clinical trial costs were estimated at $137–$206 million for Merck and $128–$192 million for GSK. The authors also estimated “manufacturing capital costs” and which resulted in total costs from $137- $539 million for Merck's RotaTeq, and $128-$392 million for GSK's Rotarix. After adjusting for inflation and capitalization at 3%, total R&D costs were estimated at $205-$644 million for Merck and $172 -$551 for GSK. All figures are expressed in 2008 USD.
Gunn et al. (2019) provided estimations of R&D costs for vaccine development at the European Vaccine Initiative (EVI), a not-for-profit organization that supports the development of vaccines against diseases of poverty and emerging infectious diseases. Preclinical stage costs for 3 vaccine candidates were estimated at EUR 2,483,333 and phase I costs were estimated at EUR 1,500,000. The type of costs included in the estimation and the period in which the costs were incurred are not specified.
Odevall et al. (2018) estimated the R&D costs to develop a cholera vaccine (euvichol) through a global public-private partnership between the International Vaccine Institute (IVI) and Eubiologics. They estimated that it took “6 years and 10 months, and a total cost of approximately 19.7 million USD (including all costs for IVI and Eubiologics)”. The paper was published in 2018, but it is not clear if costs were adjusted for inflation and seen to include only out-of-pocket costs.
Terry et al. (2018) developed a modeling tool to estimate the costs of launching new health products called the Portfolio-To-Impact (P2I) Model. The model is based on assumptions for costs, timeframes and attrition rates for each phase of development from late preclinical stage to phase III clinical trials. The assumptions were based on Parexel’s R&D cost sourcebook and refined by interviews “with a wide variety of stakeholders from Product Development Partnerships, biopharmaceutical and diagnostic companies, and major funders of global health R&D”. The model has different assumptions for different types of products, called “archetypes”, including vaccines, new chemical entities, repurposed drugs, biologics and diagnostics. The P2I Model was further refined by Young et al. (2018). A summary of the cost assumptions per each archetype as per version 2 of the P2I Model is provided in the figure below.
Source: Young et al., 2018, p. 8.
2) Drivers of R&D costs
Sertkaya et al. (2014) also analyzed the main factors impacting the costs of R&D. They conclude that “the factors that contribute the most to costs across all trial phases include Clinical Procedure Costs (15 to 22 percent), Administrative Staff Costs (11 to 29 percent), Site Monitoring Costs (nine to 14 percent), Site Retention Costs (nine to 16 percent), and Central Laboratory Costs (four to 12 percent)”.
The literature review conducted by Mestre-Ferrandiz et al. (2012) has a section dedicated to main drivers of R&D costs. The study highlights that differences in success rates and development time drive the costs up or down and also has an effected in the cost of capital. In what related to out-of-pocket costs, the main element “is the cost of clinical trials, which is affected by the cost per patient and the number of patients required to collect sufficient data.” The review suggested that clinical trial costs had increased over time, due to increased complexity. To reduce costs, a few changes in the development process occurred, such as outsourcing to clinical research organizations (CROs) and location trials emerging economies (Africa, Asia, Eastern Europe, Latin America and the Middle East), which can reduce costs “both because local costs are lower and because patient recruitment may be faster.” The author highlights, however, that “although more clinical trials are being conducted in emerging markets, especially Phase III trials, the majority of clinical trials still are conducted in the US and Western Europe, for reasons related to regulatory conditions, relevant expertise and infrastructure.” Technological challenges are also appointed as a factor driving costs up, as targeted diseases are more complex than the ones target before and “personalized/stratified medicine” have more narrow patient population.
3) Returns on R&D
Some studies have compared the revenues obtained by the selling of the drugs to their estimated R&D costs. The above-mentioned series of report on pharmaceutical R&D conducted by Deloitte Center for Health Solutions also calculates the rate of return on R&D. The last 2019 report (Deloitte, 2019) estimates the average internal rate of return (IRR) for the largest pharmaceutical companies (original cohort) to be 1.8% in 2019, down from 10.1% in 2010. For the mid-to-large companies included in the extension cohort, they estimate returns at 6.2% in 2019, down from 17.4% in 2015. They highlight lower return rates are mainly due to “assets terminations”. The figure below summarizes the estimated return rates from the reports published in the last decade.
Source: Deloitte Center for Health Solutions, 2019, p. 09.
Tay-Teo et al. (2019) investigated the return on R&D investments for cancer drugs by comparing incomes from sales to estimated R&D costs. Data from sales was extracted from public financial reports of “originator” pharmaceutical companies (defined as “companies that held patents or marketing rights”). The initial dataset included all cancer drugs approved by US FDA from 1989-2017, a total of 156, out of which 99 had enough data to be included in the analysis. The authors found that “compared with the total risk-adjusted R&D cost of $794 million (range, $2827-$219 million) per medicine estimated in the literature, by the end of 2017, the median cumulative sales income was $14.50 (range, $3.30-$55.10) per dollar invested for R&D. Median time to fully recover the maximum possible risk-adjusted cost of R&D ($2827 million) was 5 years (range, 2-10 years; n = 56)”. Figures are expressed in 2017 US dollars and were adjusted for inflation. They conclude that “cancer drugs, through high prices, have generated returns for the originator companies far in excess of possible R&D costs”.
Prasad and Mailankody (2017), in their above-mentioned paper investigating the R&D costs of cancer drugs, also calculated the total revenue from sales of the 10 drugs since approval to be USD$67.0 billion compared with total R&D spending of USD$9.1 billion (including 7% opportunity costs). The authors concluded that the revenue since approval is substantially higher than the R&D costs.
The report by the US Congress Office of Technology Assessment – OTA (1993) also calculated the average return in investment and concluded that “each new drug introduced to the U.S. market between 1981 and 1983 returned, net of taxes, at least $36 million more to its investors than was needed to pay off the R&D investment. This surplus return amounts to about 4.3 percent of the price of each drug over its product life”.
More transparency on underlying data used to estimate R&D costs
More information on the impact of tax credits and regulatory incentives that might reduce net costs of R&D borne by product developer
More studies on R&D costs from not-for-profit product development organizations (e.g. public institutions, academia and product development partnerships)
Studies targeted for specific informational needs, i.e. studies responding to the analytical needs of company executives or investors may not adequately respond to the needs of policymakers, buyers, or the public.
Costs estimations for the development of other health products beyond drugs (e.g. vaccines, diagnostics, other medical devices)
 There is a separate research synthesis on R&D timeframes and attrition/success rates.
 The author notes that “some papers have estimated the cost of developing new drugs under the auspices of what have become known as “product-development partnerships” (PDPs), especially for neglected diseases. The discussion of these estimates is beyond the scope of this publication” (p.8).
 Definition provided at the glossary of the study (p. 79), “CMRI: CMRI, acquired by Thomson Reuters in 2006, began researching issues in R&D in the early 1980s as the Centre for Medicines Research (CMR). It maintains various databases of drugs/biologics and biopharmaceutical industry activities.”
 Wiggins SN. The cost of developing a new drug. Washington, DC: Pharmaceutical Manufacturers Association; 1987. Apud Mestre-Ferrandiz et al. 2012.
 Hansen, R.W. (1979) The pharmaceutical development process: Estimates of current development costs and times and the effects of regulatory changes. In: R.I. Chien (ed.) Issues in Pharmaceutical Economics. Lexington, MA: Lexington Books, pp. 151–187. Apud Mestre-Ferrandiz et al. 2012.
* For the purposes of this review, we have established three categories to describe the state of the literature: thin, considerable, and rich.
- Thin: There are relatively few papers and/or there are not many recent papers and/or there are clear gaps
- Considerable: There are several papers and/or there are a handful of recent papers and/or there are some clear gaps
- Rich: There is a wealth of papers on the topic and/or papers continue to be published that address this issue area and/or there are less obvious gaps
Scope: While many of these issues can touch a variety of sectors, this review focuses on medicines. The term medicines is used to cover the category of health technologies, including drugs, biologics (including vaccines), and diagnostic devices.
Disclaimer: The research syntheses aim to provide a concise, comprehensive overview of the current state of research on a specific topic. They seek to cover the main studies in the academic and grey literature, but are not systematic reviews capturing all published studies on a topic. As with any research synthesis, they also reflect the judgments of the researchers. The length and detail vary by topic. Each synthesis will undergo open peer review, and be updated periodically based on feedback received on important missing studies and/or new research. Selected topics focus on national and international-level policies, while recognizing that other determinants of access operate at sub-national level. Work is ongoing on additional topics. We welcome suggestions on the current syntheses and/or on new topics to cover.