Financial management of large, multi-center trials in a challenging funding milieu

Financial management of large, multi-center trials in a challenging funding milieu Background: Randomized clinical trials that have public health implications but no or low potential for commercial gain are predominantly funded by governmental (e.g., National Institutes of Health (NIH)) and not-for-profit organizations. Our objective was to develop an alternative clinical trial site funding model for judicious allocation of declining public research funds. Methods: In the Vitamin D and Type 2 Diabetes (D2d) study, an NIH-supported, large clinical trial testing the effect of vitamin D supplementation on incident diabetes in 2423 participants at high risk for diabetes, a hybrid financial management model for supporting collaborating clinical sites was developed and applied. The funding model employed two reimbursement components: Core (for study start-up and partial efforts throughout the study, ~40% of the total site budget), invoiced by sites, and Performance-Based Payments (for successful enrollment of participants and completion of follow-up visits, ~60% of the total site budget), automatically issued to the sites by the Coordinating Center based on actual recruitment and visits conducted. Underperforming sites transitioned to Performance-Based Payments only. Results: Recruitment occurred from October 2013 through December 2016, requiring one additional year than the 2-year projection. Median enrollment at each site was 88 participants (range 29–318; 20 to 205% of the site target). At the end of year 1, study-wide recruitment was at 12% of the target (vs. 50% projected) and 12% of the total grant award was invested. The model constantly evaluated sites’ needs and re-allocated resources to meet the study enrollment goal. If D2d had issued cost reimbursement subaward agreements and sites invoiced for their entire budget, 83% of the award would have been spent for all study activities over the first 4 years of the trial compared to 65% of the award spent (US$26M) under the hybrid model used by D2d. Conclusions: It is feasible to foster a hybrid financial management approach to steward limited available public funds for research in a dynamic and consistent way that does not compromise the trial’s scientific integrity and ensures conservation of funds to complete recruitment and continue to follow up participants. Keywords: Clinical trial, Diabetes, Budget, Hybrid, Financial management, National Institutes of Health Background potential for financial gain (e.g., study of rare diseases, Randomized clinical trials are at the top of the hierarchy non-patentable therapies, dietary supplements) are pre- of research designs and provide the highest level of dominantly funded by not-for-profit or governmental evidence to advance clinical care. One of the major bar- agencies, the largest of which in the USA is the National riers to undertaking clinical trials is expense. In 2013, Institutes of Health (NIH). Although the importance of the estimated average per-participant cost in industry- clinical trials is well recognized by the NIH [2], funding sponsored trials was US$36,500 [1]. Trials that address is becoming increasingly challenging. From 1994 to clinically important questions but have no or low 2003, the NIH budget more than doubled from US$11. 0B to US$27.1B with a further gradual increase until * Correspondence: olovegreen@tuftsmedicalcenter.org 2010 when the NIH budget started to decline [3]. How- Olivia Lovegreen and Danielle Riggs contributed equally to this work. ever, after accounting for inflation, the NIH budget has Research Administration, Department of Medicine, Tufts Medical Center, 800 declined every year since 2003 and recent proposals by Washington St, Boston, MA 02111, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Lovegreen et al. Trials (2018) 19:267 Page 2 of 8 the White House recommend roughly 18% reduction in D2d has been published [6]. The study is approved and the NIH budget [4]. Therefore, alternate methods aiming monitored by an independent data and safety monitoring to reduce the costs of large randomized trials, supported board and the institutional review board of each collabor- by the NIH and other governmental agencies and not- ating clinical research site. Target participants were adults for-profit organizations, are needed. at high risk for diabetes, defined as meeting at least two of In large, multi-center clinical trials, the proportion that three glycemic criteria for prediabetes established by the is allotted to collaborating clinical sites to recruit and American Diabetes Association in 2010 [7]. Following retain participants is the largest and most variable com- randomization, participants were seen twice a year and ponent of the budget. Large, multi-center clinical trials had laboratory testing for ascertainment of diabetes. have historically employed two models to manage clin- D2d is being conducted at 22 US collaborating clinical ical site budgets. The cost reimbursement model pro- sites (www.d2dstudy.org/sites) selected by the D2d vides subaward agreements that cover amounts for Coordinating Center and the funding agency because of personnel effort (i.e., site investigators and staff) and their ability to recruit and retain a diverse population of participant expenses (e.g., recruitment, testing, follow-up people at risk for diabetes. Potential sites submitted ap- visits). Under this approach, the subaward amounts are plications that detailed past research experience with pre-determined based on the site’s projected enrollment emphasis on participant recruitment and retention, and and a site could be reimbursed for all expenses despite provided information on access to potential participants enrolling few participants. Under this model, if efforts and a D2d-specific recruitment plan. Applications were are not promptly undertaken to reduce the costs of reviewed by the planning committee and objectively underperforming sites, a large proportion of the grant scored. The site budget was not a factor in the selection award could be depleted prior to the trial reaching the process. Each site was expected to enroll approximately targeted enrollment, reducing funds for the remaining 100 to 150 participants towards the study-wide enroll- trial activities. The second model (“performance-based ment target of 2382. D2d is an event-driven trial that only”) provides a pre-determined reimbursement amount will continue until the required number of diabetes out- per participant enrolled. This approach conserves funds come events is reached, which is expected to occur in during slow recruitment and is results-oriented (e.g., fo- late 2018. cusing on enrollment metrics) but requires sites to “front” the costs for study start-up and initial recruitment efforts. Overview of the financial management model Additionally, this model may not fully consider the vari- The hybrid financial management model utilized by D2d ability in regional costs that make research more expen- has two reimbursement components for the sites, Core sive to conduct in certain high-cost areas [5]. Under this and Performance-Based Payments (PBPs). Core is pro- model, the risk of non-performance rests with the clinical vided primarily for study start-up (e.g., institutional site and the amount of anticipated payments is unknown, review board (IRB) submission, staff training) and partial which makes recipient institutions apprehensive [5]. effort throughout the study for personnel-based activities In the setting of declining public budgets for research, (recruitment planning, pre-screening, participation in many groups within the USA are exploring alternative committees). Performance-Based Payments are provided approaches to reduce the costs of clinical trials [1]. We quarterly for successful enrollment of participants and describe a hybrid financial management model for col- completion of follow-up visits in the preceding 3 months. laborating clinical sites, which we applied in the Vitamin D and Type 2 Diabetes (D2d) study, an NIH-supported, Budget year 1: determination of Core vs. PBP value large clinical trial testing the safety and efficacy of vita- At the time of the grant application, each site developed a min D supplementation on the development of diabetes standard 5-year NIH budget based on the targeted number in people at high risk for diabetes [6]. of enrolled (i.e., randomized) participants at the site. The D2d Coordinating Center provided guidance on personnel Methods effort for a typical site (Table 1). Each site’s approved 5-year Overview of the study design budget was used to determine site-specific Core and PBP D2d is a US-based, multi-center, randomized (1:1), double- amounts. At the start of the study, Core was defined as blind, placebo-controlled, parallel-group, primary preven- 50% of annual personnel expenses. Each site invoiced D2d tion clinical trial comparing oral administration of 4000 IU/ quarterly for Core support; the remaining 50% of personnel day of cholecalciferol (vitamin D ) vs. placebo in people at effort and all non-personnel expenses were recovered high risk for diabetes who are followed for incident diabetes through PBPs. PBPs were calculated by deducting the pro- (primary outcome) for approximately 3 years after jected value of the Core (over the 5 years) from the ap- randomization (www.d2dstudy.org). Cancer and cardiovas- proved 5-year budget and dividing the balance by the cular disease are key secondary outcomes. The design of projected site enrollment to establish a site-specific PBP Lovegreen et al. Trials (2018) 19:267 Page 3 of 8 Table 1 Suggested breakdown by the D2d Coordinating Center Table 3 Scheme for Performance-Based Payments per participant of personnel effort for a typical study site that is projected to visit for a typical D2d site that is projected to enroll (i.e., randomize) enroll (i.e., randomize) and follow 125 participants and follow 125 participants Year 1 Year 2 Year 3 Year 4 Year 5 Per participant visit reimbursement (%) (%) (%) (%) (%) Visit % of total Performance-Based US$ Principal investigator 20 20 20 20 20 Payment amount and co-investigator(s) Month 00 20 982 (randomization) Research coordinator 100 100 100 75 75 Month 6 6 295 Research assistant 50 100 100 75 75 Assumptions: 125 participants will be randomized by year 2. The bulk of participant Month 12 20 982 follow-up is expected to occur in years 2–5 Month18 6 295 Month 24 20 982 amount per enrolled participant (Table 2). The total PBP Month 30 6 295 amount per enrolled participant was then distributed Month 36 22 1081 across seven scheduled study visits, proportionally to the Total 100 4912 work involved at each visit (Table 3). The Coordinating Month 42 6 295 Center issued quarterly PBPs to the sites based on actual visits conducted, verified by data in the study’selectronic Month 48 22 1081 data capture system. Based on this model, a site would only Total site-specific Performance-Based Payment amount per enrolled participant who completes all seven scheduled study visits, month 00 to month 36 (see also receive its total approved 5-year budget if the targeted text and Table 2). While the actual amount paid per participant visit varied by site number of participants were enrolled and followed for the (column 3), the percent of the total Performance-Based Payment amount allocated in each visit (column 2) was identical for all sites study duration. Visits beyond month 36 (month 42, etc.) are not included in the total site-specific Performance-Based Payment calculations becausetheaverageparticipant is expected to be followed for 36 months. A site that enrolls participants early in the recruitment period and follows these participants beyond month 36 will be Budget year 2: revisions to the model/addition of new reimbursed for additional visits (e.g., month 42 and month 48) in the same sites way as other similar visits (e.g., month 30 and month 36) The model was revised in year 2 to better accommodate the sites’ evolving needs. The central components (Core and PBPs) were preserved, but Core was redefined as 40% Budget year 3 and beyond: further revisions to the model of the total annual expenses (i.e., personnel effort and Because recruitment continued through year 3 and non-personnel costs) (Table 2). Quarterly PBPs continued part of year 4, the model was further revised. In years to be made to the sites based on successful enrollment of 3 and 4, sites received the same Core (in absolute US$) participants and completion of follow-up visits. as in year 2 (Table 2). The model reduced the planned Due to a delayed start of recruitment and lag in meeting Core in year 5, such that the total Core received over enrollment targets, D2d added four new sites in year 2. 5 years was the same as the original proposed Core The new sites submitted a standard 4-year NIH budget (Table 2). Sites were reminded that as the study pro- (grant years 2 to 5), and the revised methodology gressed they would receive a larger proportion of their described above was used to determine Core and PBPs. budget via PBPs. Table 2 Fundamental scheme of the hybrid financial management model utilized by D2d for a typical site that projected to enroll (i.e., randomize) 125 participants Year 1 Year 2 Year 3 Year 4 Year 5 Total Personnel (effort and fringe benefits) 178,135 183,100 183,100 137,171 137,171 Non-personnel 44,534 65,586 60,208 22,043 12,297 Total proposed budget 222,669 248,686 243,308 159,214 149,468 1,023,345 Core Proposed Core 89,068 91,550 91,550 68,586 68,586 409,339 Actual Core 89,068 99,474 99,474 99,474 21,848 409,339 Performance-Based Payments PBP amount per enrolled participant: 4912 614,007 Amounts shown in US$; numbers include directsand 53%indirectrate. Theschemeisflexibleboth from year to year and from site to site, while maintaining consistency Proposed Core was defined at the start of the study as 50% of annual personnel costs Actual Core varied from year to year (see text) while the total amount over 5 years did not exceed the total proposed Core (i.e., US$409,339) Lovegreen et al. Trials (2018) 19:267 Page 4 of 8 Transition to PBP-only for underperforming sites US$7.4M would have been needed to complete recruitment The model implemented a uniform approach to manage (Fig. 2). If D2d had issued cost reimbursement subaward underperforming sites. If the D2d leadership determined agreements and sites invoiced for their entire budget, 83% that there was potential for recruitment to improve at of the award (US$33M) would have been spent compared an underperforming site, the Coordinating Center con- to 65% of the award spent (US$26M) under the hybrid tinued to provide Core support for a limited “grace model used by D2d over the first 4 years of the trial (Fig. 2). period” of 3, 4, or 6 months. During that time, the Coordinating Center worked closely with the underper- forming site to establish new, realistic enrollment goals Budget year 1: determination of Core vs. PBP value and supported the site to achieve the revised goals. In year 1, individual site budgets varied considerably If a site did not meet its new target, site payments (range US$3324 to US$8406 per participant to recruit transitioned to PBP-only. The PBP-only alternative elim- and follow), largely due to varying salaries for research inated Core but increased the PBP amount per enrolled staff and fringe benefits (range US$38,027 to US$82,000 participant accordingly, so that the site was not penal- for research coordinators; US$25,094 to US$45,000 for ized for continuing to enroll and follow participants. research assistants; 15 to 41% for fringe benefits). These differences were attributed to regional differences (i.e., Primary outcome New York City vs. smaller cities in the Midwest), staff The primary outcome of interest in relation to the seniority, and expertise level. hybrid model used by the D2d study was total expenses In year 1, site activation and participant recruitment under this model compared to total expenses under a were slower than anticipated, which had a trickle-down traditional model. effect on the model. Sites invoiced Core for 50% of actual personnel expenses, but could not fully recoup Results the remaining 50% through PBPs due to slow recruit- Recruitment occurred from October 2013 through ment. At the end of year 1, study-wide recruitment was December 2016, and 2423 participants were randomized. at 12% of the target (vs. 50% projected) and 12% of the The D2d clinical sites were located throughout the USA total grant award was invested. Of the amount invested (Fig. 1) and included academic non-profit private med- in year 1, 54% went to the clinical sites (46% as Core ical centers, public medical centers, or Veterans and 8% as PBPs) and 46% went to other components Administration medical centers. (Coordinating Center, central laboratory, and drug The hybrid model used by D2d worked well to complete distribution). During year 1, two sites were discontin- recruitment and have funds remaining to follow partici- ued, one due to personnel turnover and one for signifi- pants. Under a cost reimbursement model, an additional cant underperformance. Fig. 1 Location of collaborating clinical sites in the D2d study Lovegreen et al. Trials (2018) 19:267 Page 5 of 8 Fig. 2 Comparison of the hybrid model used by D2d vs. a cost reimbursement subaward agreement for clinical sites. The cost reimbursement model includes the originally submitted budgets for sites (as cost reimbursement agreements) and actual expenses for non-site components (Coordinating Center, drug distribution, central lab) through the end of the recruitment phase without any budget reductions for underperformance. The hybrid model used by D2d includes actual spending (clinical sites and non-site components) through the end of the recruitment phase Budget year 2: revisions to the model/addition of new sites Budget year 3 and beyond: further revisions to the model Revisions to the model and addition of new sites in year As recruitment continued beyond year 2, into years 3 2 were critical to the completion of recruitment. Revi- and 4, the changes in Core payments were critical to sions to the Core gave sites the flexibility to allocate maintain recruitment activities. Increasing Core pay- Core support towards any study-related expense to meet ments in years 3 and 4 and reducing Core payments in site goals. At the end of recruitment, half of the sites year 5 allowed recruitment to be completed within the met or exceeded the 80% mark of their recruitment tar- original budget. get (Fig. 3). As in many large, multi-center trials, there was a wide range of recruitment success. Median enroll- Transition to PBP-only for underperforming sites ment at each site was 88 participants (range 29–318; 20 Transitioning underperforming sites to PBP-only agree- to 205% of the site target; Fig. 3). Three out of four new ments did not have an appreciable effect on recruitment sites (75%) met or exceeded their enrollment target, vs. or retention. Sites that transitioned to PBP-only agree- four out of 18 (22%) for the original sites. Of the D2d ments enrolled an average of 1.05 participant per month cohort, 20% was accounted for by the four new sites. before the transition, compared to 1.35 participant per Fig. 3 Enrollment in the D2d study. Enrollment is shown as percent of target, study-wide (dark blue bar) and for each clinical site (light blue bars). Horizontal line indicates the 80% mark of the enrollment target. Number of randomized participants is shown at the top of each bar. (√) indicates new sites added in year 2. (*) indicates sites that transitioned to a Performance-Based Payment-only arrangement Lovegreen et al. Trials (2018) 19:267 Page 6 of 8 month after the transition. Sites that continued to re- electronic health record (EHR)-based recruitment ap- ceive Core and PBP randomized 2.5 participants per proach, especially for a trial with laboratory-based inclusion month. Sites receiving PBP-only completed 92% of all criteria. To allocate resources based on effort and results, participant encounters vs. 95% for sites receiving Core the model established expectations in terms of recruitment and PBPs. and retention targets and timelines. Site recruitment was monitored weekly by the study leadership, and the model encouraged sites to abandon inefficient approaches. The Discussion bulk of the funds provided to the sites matched recruitment The hybrid financial management model used in the D2d and retention results, especially as the study progressed. study was designed to promote judicious investment of At the outset, recognizing that few sites would reach their precious resources in a way that does not compromise goal, the Coordinating Center estimated that the average recruitment and follow-up of participants. The fundamen- site would enroll about 80% of its projected target, while a tal principle behind the model is to allocate the total few sites would exceed their target. The prediction was cor- award earmarked for the sites proportional to their effort rect, as exactly half of the sites met or exceeded the 80% and results, while maintaining open-mindedness and flexi- mark at the end of the recruitment period. However, the bility to meet sites’ evolving needs. The methodology used range of recruitment success (20 to 205% of the stated goal) to determine the Core and PBPs based on each site’s was larger than anticipated. Therefore, flexibility to divert budget of what they expected their actual study costs over funds away from underperforming sites to match needs of 5 years was novel. In the process, we learned several well-performing sites was a key component of the model. lessons (Table 4)that we discuss below. Maintain open-mindedness and flexibility when Encourage multi-disciplinary teamwork addressing sites’ concerns In D2d, the study’s Principal Investigator, Coordinating Although the hybrid payment model resulted in a more Center, Research Administration at the prime grantee efficient investment of research funds in direct propor- institution, and the funding agency worked collabora- tion to results, many sites faced budget deficiencies due tively to design and employ a principles-based financial to underperformance and unanticipated issues, including management model to ensure appropriate stewardship increases in fringe rate, facilities and administrative rate, of the available funds. The model aimed for consistency and other expenses not accounted for in their original while recognizing sites’ unique circumstances. At least budget. Sites also did not fully appreciate that they in part, the hybrid model used by D2d worked well would receive their entire approved 5-year budget only if because people from different disciplines established a they fully met their enrollment and retention goals. For close and dynamic collaboration and decisions were example, if a site reached 80% of its enrollment target made in a cooperative fashion with the clinical sites. and successfully completed all follow-up visits, its PBPs would be reduced proportionally by 20%. Allocate resources based on results, not just effort Recognizing sites’ unique circumstance is important in Although effort and results often correlate, the association large, multi-center trials and decisions need to be made in is far from perfect, as effort may be exerted inefficiently or a consistent manner. The D2d leadership kept an open ineffectively. For example, bulk mailing of non-targeted let- mind throughout the recruitment phase—the most variable ters is an inefficient recruitment method compared to an phase of a clinical trial—and worked individually with sites, recognizing the uniqueness of the challenges faced by each, and strived to be flexible. For example, in each year during Table 4 The D2d hybrid budget management model: lessons the recruitment phase, the D2d leadership used a uniform learned approach to adjust the Core to better accommodate sites’ � Encourage multi-disciplinary teamwork that engages the Principal needs. For outlier sites, the D2d leadership re-established Investigator, Coordinating Center, Research Administration at the expectations and adjusted the budget as needed. When prime, subaward institutions, and the funding agency. faced with the potential of a site withdrawing from the � Allocate resources based on results, not just effort. study due to financial issues, the D2d leadership worked � Maintain open-mindedness and flexibility when addressing sites’ with the site to modify recruitment goals and its budget, concerns. while staying within the framework of the model. � Provide Core support and sufficient resources for study start-up. � Accept that sites will be discontinued and sites will be added. Provide Core support and sufficient resources for study � Project that recruitment will take longer and the outcome event rate start-up may be lower than predicted. A key concept of the hybrid model used by D2d is to pro- � Reinforce the model’s principles to site staff. vide Core funds for study start-up (e.g., IRB submission, Lovegreen et al. Trials (2018) 19:267 Page 7 of 8 study setup) and partial salary support (e.g., pre-screening [8], which may require study extension, thereby adding of participants, participation in committees). While a strain to the budget. The hybrid model used by D2d con- PBP-only model is conservative and advantageous to the served funds to ensure completion of recruitment while total study budget, under such model, sites may be reluc- conserving funds to continue following participants. If tant to invest in the initial effort, including exploring in- D2d had issued cost reimbursement agreements and sites novative recruitment methods. Study-wide, providing invoiced for their entire budget without any modifications Core funds may be a risky approach if many sites under- for underperformance, nearly the entire grant award perform. In D2d, providing Core for a proportion of an- would have been spent at the completion of recruitment. nual personnel effort only did not work. Redefining the Core as 40% of the annual total budget (including both Reinforce the model’s principles to site staff personnel and non-personnel) worked well and allowed A hybrid model can be new to investigators and research sites to overcome administrative concerns. The Core com- administration staff; therefore, it is important to reinforce ponent proved to be critical to the success of D2d as it the principles of the model throughout the study. When allowed sites to invest in the successful launch of the trial. executing the original subawards, the Coordinating Center provided a written document that described the model. As Accept that sites will be discontinued and sites will be the model was enhanced each year, the D2d leadership added maintained open communication with sites to optimize Despite extensive experience and initial enthusiasm, a few understanding and application of the model, especially sites will never get off the ground and many will underper- given turnover of site research administrative staff. form. There are many reasons that can prevent sites from reaching their full potential, including natural disasters, Limitations research staff relocation or retirement, illness affecting key The model worked for D2d, which is a focused and personnel, and shifting priorities at the site. It is important highly structured trial with laboratory-based eligibility for the leadership team to identify sites that have the criteria and outcomes. The model may be more difficult potential to improve with appropriate guidance and sites to adapt in a more complicated trial with multiple out- that need to be discontinued. The decision to discontinue comes. Other groups are exploring different methods to a site is exceedingly difficult because of financial, human minimize costs of clinical trials, such as use of mobile (personnel), academic (collegial), and emotional implica- technology or electronic medical records [9, 10]. Add- tions. There are ethical concerns in relation to the investi- itionally, the model was applied in a trial supported by gators’ duty to enrolled participants. However, timely the NIH, and results may not apply to non-NIH trials or separation is necessary as delay only utilizes more funds those outside of the USA. and makes the discontinuation even more problematic, es- pecially when many participants need to be administra- Conclusions tively withdrawn. Discontinuation of sites needs to follow The NIH-supported D2d clinical trial employed a hybrid a consistent process, and the decision needs to be unani- financial management model, comprised of two compo- mous by the study leadership. nents (Core and Performance-Based Payments) to man- It is certain that not all original sites will perform as age the study’s budget in a dynamic and creative way to projected; therefore, adding new sites should be antici- effectively invest finite resources. The development and pated and planned. Adding sites while the trial is employment of the model is the result of a cooperative ongoing is logistically challenging; however, the leader- effort between Research Administration at the prime ship, based on accumulated study-specific experience, is grantee institution, Principal Investigator, Coordinating better prepared to evaluate and select new sites whose Center, and the primary funding agency. Anticipating a strengths match the study’s needs. The addition of new delay in recruitment, the model continuously evaluated sites proved critical to the success of D2d. sites’ needs and re-allocated resources to meet the study enrollment goal and succeeded in managing the study’s Recruitment will take longer and the outcome event rate funds during the extended recruitment period in a way may be lower than projected that did not compromise the trial’s scientific integrity. It is rare for large trials to complete recruitment within Abbreviations the original timeline. In D2d, the recruitment period was D2d: Vitamin D and Type 2 Diabetes; NIDDK: National Institute of Diabetes projected to last 2 years; however, despite the sites’ best and Digestive and Kidney Diseases; NIH: National Institutes of Health; PBPs: Performance-Based Payments efforts, including adoption of a very successful electronic health record-based recruitment approach, enrollment Acknowledgements was slower than projected. Furthermore, the rate of occur- We thank the D2d investigators, staff, and trial participants for their rence of the primary outcome can be lower than projected outstanding dedication and commitment to the study. Lovegreen et al. Trials (2018) 19:267 Page 8 of 8 Funding 4. Reardon S, Tollefson J, Witze A, Ross E. US science agencies face deep cuts The planning phase of D2d was funded by the National Institute of Diabetes in Trump budget. 2017. http://www.nature.com/news/us-science-agencies- and Digestive and Kidney Diseases (NIDDK) through a multi-center clinical study face-deep-cuts-in-trump-budget-1.21652. Accessed 7 Apr 2017. implementation planning grant (U34) to Tufts Medical Center in Boston, MA 5. Larson GS, Carey C, Grarup J, Hudson F, Sachi K, Vjecha MJ, Gordin F, Group (U34DK091958; principal investigator A.G.P.). Planning was also supported in I. Lessons learned: infrastructure development and financial management part by the Intramural Research Program of the NIDDK. The conduct of D2d is for large, publicly funded, international trials. Clin Trials. 2016;13:127–36. primarily supported by the NIDDK and the Office of Dietary Supplements of the 6. Pittas AG, Dawson-Hughes B, Sheehan PR, Rosen CJ, Ware JH, Knowler National Institutes of Health through a multi-center clinical study cooperative WC, Staten MA, Group DdR. Rationale and design of the Vitamin D and agreement (U01DK098245; principal investigator A.G.P.) to Tufts Medical Center, Type 2 Diabetes (D2d) study: a diabetes prevention trial. Diabetes Care. where the D2d Coordinating Center is based. The U01 grant mechanism 2014;37:3227–34. establishes the NIDDK project scientist (M.A.S.) as a member of the D2d 7. American Diabetes Association. Standards of medical care in diabetes–2010. Research Group. The study also received secondary funding from the Diabetes Care. 2010;33(Suppl 1):S11–61. American Diabetes Association (1-14-D2d-01). Educational materials are 8. Lappe J, Watson P, Travers-Gustafson D, Recker R, Garland C, Gorham E, provided by the National Diabetes Education Program. No pharmaceutical Baggerly K, McDonnell SL. Effect of vitamin D and calcium supplementation manufacturers contributed to the planning, design, or conduct of D2d. on cancer incidence in older women: a randomized clinical trial. JAMA. Study pills were purchased from an independent nutritional supplement 2017;317:1234–43. manufacturing company that has no association with any members of the 9. Eapen ZJ, Lauer MS, Temple RJ. The imperative of overcoming barriers to D2d Research Group. The D2d investigators and the NIDDK project scientist the conduct of large, simple trials. JAMA. 2014;311:1397–8. were responsible for the design and conduct of the study; collection, 10. Clinical Trials Transformation Initiative. Mobile clinical trials. https://www.ctti- management, analysis, and interpretation of the data; preparation, review, clinicaltrials.org/programs/mobile-clinical-trials. and approval of the manuscript; and decision to submit the manuscript for publication. Availability of data and materials This study is registered on ClinicalTrials.gov (NCT01942694). Data are not publically available but are available from the corresponding author on reasonable request and in accord with NIH regulations. Authors’ contributions All authors contributed to the design of the D2d financial management model. OL, DR, MAS, and AGP contributed to the execution of the model. All authors read and approved the final manuscript. Ethics approval and consent to participate The study is approved and monitored by an independent data and safety monitoring board (appointed by the National Institute of Diabetes and Digestive and Kidney Diseases) and the institutional review board of each collaborating clinical research site. Written informed consent was obtained from each participant prior to study procedures. Competing interests The authors declare that they have no competing interests. Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Author details Research Administration, Department of Medicine, Tufts Medical Center, 800 Washington St, Boston, MA 02111, USA. Kelly Government Solutions for National Institute of Diabetes and Digestive and Kidney Diseases, 900 Rockville Pike, Bethesda, MD 20892, USA. Division of Endocrinology, Diabetes and Metabolism, Department of Medicine, Tufts Medical Center, 800 Washington St, Boston, MA 02111, USA. Received: 7 August 2017 Accepted: 10 April 2018 References 1. Battelle Technology Partnership Practice. Biopharmaceutical industry- sponsored clinical trials: impact on state economies. 2015. http://phrma-docs. phrma.org/sites/default/files/pdf/biopharmaceutical-industry-sponsored- clinical-trials-impact-on-state-economies.pdf. 2. Hudson KL, Lauer MS, Collins FS. Toward a new era of trust and transparency in clinical trials. JAMA. 2016;316:1353–4. 3. Johnson J. Brief history of NIH funding: fact sheet. Congressional Research Service. Prepared for Members and Committes of Congress. 2013. https:// www.everycrsreport.com/files/20131223_R43341_73c480e1172aef191e8cb81 fc7ecbf895ad830b4.pdf. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Trials Springer Journals

Financial management of large, multi-center trials in a challenging funding milieu

Free
8 pages
Loading next page...
 
/lp/springer_journal/financial-management-of-large-multi-center-trials-in-a-challenging-MoF0yJ2evO
Publisher
BioMed Central
Copyright
Copyright © 2018 by The Author(s).
Subject
Medicine & Public Health; Medicine/Public Health, general; Biomedicine, general; Statistics for Life Sciences, Medicine, Health Sciences
eISSN
1745-6215
D.O.I.
10.1186/s13063-018-2638-z
Publisher site
See Article on Publisher Site

Abstract

Background: Randomized clinical trials that have public health implications but no or low potential for commercial gain are predominantly funded by governmental (e.g., National Institutes of Health (NIH)) and not-for-profit organizations. Our objective was to develop an alternative clinical trial site funding model for judicious allocation of declining public research funds. Methods: In the Vitamin D and Type 2 Diabetes (D2d) study, an NIH-supported, large clinical trial testing the effect of vitamin D supplementation on incident diabetes in 2423 participants at high risk for diabetes, a hybrid financial management model for supporting collaborating clinical sites was developed and applied. The funding model employed two reimbursement components: Core (for study start-up and partial efforts throughout the study, ~40% of the total site budget), invoiced by sites, and Performance-Based Payments (for successful enrollment of participants and completion of follow-up visits, ~60% of the total site budget), automatically issued to the sites by the Coordinating Center based on actual recruitment and visits conducted. Underperforming sites transitioned to Performance-Based Payments only. Results: Recruitment occurred from October 2013 through December 2016, requiring one additional year than the 2-year projection. Median enrollment at each site was 88 participants (range 29–318; 20 to 205% of the site target). At the end of year 1, study-wide recruitment was at 12% of the target (vs. 50% projected) and 12% of the total grant award was invested. The model constantly evaluated sites’ needs and re-allocated resources to meet the study enrollment goal. If D2d had issued cost reimbursement subaward agreements and sites invoiced for their entire budget, 83% of the award would have been spent for all study activities over the first 4 years of the trial compared to 65% of the award spent (US$26M) under the hybrid model used by D2d. Conclusions: It is feasible to foster a hybrid financial management approach to steward limited available public funds for research in a dynamic and consistent way that does not compromise the trial’s scientific integrity and ensures conservation of funds to complete recruitment and continue to follow up participants. Keywords: Clinical trial, Diabetes, Budget, Hybrid, Financial management, National Institutes of Health Background potential for financial gain (e.g., study of rare diseases, Randomized clinical trials are at the top of the hierarchy non-patentable therapies, dietary supplements) are pre- of research designs and provide the highest level of dominantly funded by not-for-profit or governmental evidence to advance clinical care. One of the major bar- agencies, the largest of which in the USA is the National riers to undertaking clinical trials is expense. In 2013, Institutes of Health (NIH). Although the importance of the estimated average per-participant cost in industry- clinical trials is well recognized by the NIH [2], funding sponsored trials was US$36,500 [1]. Trials that address is becoming increasingly challenging. From 1994 to clinically important questions but have no or low 2003, the NIH budget more than doubled from US$11. 0B to US$27.1B with a further gradual increase until * Correspondence: olovegreen@tuftsmedicalcenter.org 2010 when the NIH budget started to decline [3]. How- Olivia Lovegreen and Danielle Riggs contributed equally to this work. ever, after accounting for inflation, the NIH budget has Research Administration, Department of Medicine, Tufts Medical Center, 800 declined every year since 2003 and recent proposals by Washington St, Boston, MA 02111, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Lovegreen et al. Trials (2018) 19:267 Page 2 of 8 the White House recommend roughly 18% reduction in D2d has been published [6]. The study is approved and the NIH budget [4]. Therefore, alternate methods aiming monitored by an independent data and safety monitoring to reduce the costs of large randomized trials, supported board and the institutional review board of each collabor- by the NIH and other governmental agencies and not- ating clinical research site. Target participants were adults for-profit organizations, are needed. at high risk for diabetes, defined as meeting at least two of In large, multi-center clinical trials, the proportion that three glycemic criteria for prediabetes established by the is allotted to collaborating clinical sites to recruit and American Diabetes Association in 2010 [7]. Following retain participants is the largest and most variable com- randomization, participants were seen twice a year and ponent of the budget. Large, multi-center clinical trials had laboratory testing for ascertainment of diabetes. have historically employed two models to manage clin- D2d is being conducted at 22 US collaborating clinical ical site budgets. The cost reimbursement model pro- sites (www.d2dstudy.org/sites) selected by the D2d vides subaward agreements that cover amounts for Coordinating Center and the funding agency because of personnel effort (i.e., site investigators and staff) and their ability to recruit and retain a diverse population of participant expenses (e.g., recruitment, testing, follow-up people at risk for diabetes. Potential sites submitted ap- visits). Under this approach, the subaward amounts are plications that detailed past research experience with pre-determined based on the site’s projected enrollment emphasis on participant recruitment and retention, and and a site could be reimbursed for all expenses despite provided information on access to potential participants enrolling few participants. Under this model, if efforts and a D2d-specific recruitment plan. Applications were are not promptly undertaken to reduce the costs of reviewed by the planning committee and objectively underperforming sites, a large proportion of the grant scored. The site budget was not a factor in the selection award could be depleted prior to the trial reaching the process. Each site was expected to enroll approximately targeted enrollment, reducing funds for the remaining 100 to 150 participants towards the study-wide enroll- trial activities. The second model (“performance-based ment target of 2382. D2d is an event-driven trial that only”) provides a pre-determined reimbursement amount will continue until the required number of diabetes out- per participant enrolled. This approach conserves funds come events is reached, which is expected to occur in during slow recruitment and is results-oriented (e.g., fo- late 2018. cusing on enrollment metrics) but requires sites to “front” the costs for study start-up and initial recruitment efforts. Overview of the financial management model Additionally, this model may not fully consider the vari- The hybrid financial management model utilized by D2d ability in regional costs that make research more expen- has two reimbursement components for the sites, Core sive to conduct in certain high-cost areas [5]. Under this and Performance-Based Payments (PBPs). Core is pro- model, the risk of non-performance rests with the clinical vided primarily for study start-up (e.g., institutional site and the amount of anticipated payments is unknown, review board (IRB) submission, staff training) and partial which makes recipient institutions apprehensive [5]. effort throughout the study for personnel-based activities In the setting of declining public budgets for research, (recruitment planning, pre-screening, participation in many groups within the USA are exploring alternative committees). Performance-Based Payments are provided approaches to reduce the costs of clinical trials [1]. We quarterly for successful enrollment of participants and describe a hybrid financial management model for col- completion of follow-up visits in the preceding 3 months. laborating clinical sites, which we applied in the Vitamin D and Type 2 Diabetes (D2d) study, an NIH-supported, Budget year 1: determination of Core vs. PBP value large clinical trial testing the safety and efficacy of vita- At the time of the grant application, each site developed a min D supplementation on the development of diabetes standard 5-year NIH budget based on the targeted number in people at high risk for diabetes [6]. of enrolled (i.e., randomized) participants at the site. The D2d Coordinating Center provided guidance on personnel Methods effort for a typical site (Table 1). Each site’s approved 5-year Overview of the study design budget was used to determine site-specific Core and PBP D2d is a US-based, multi-center, randomized (1:1), double- amounts. At the start of the study, Core was defined as blind, placebo-controlled, parallel-group, primary preven- 50% of annual personnel expenses. Each site invoiced D2d tion clinical trial comparing oral administration of 4000 IU/ quarterly for Core support; the remaining 50% of personnel day of cholecalciferol (vitamin D ) vs. placebo in people at effort and all non-personnel expenses were recovered high risk for diabetes who are followed for incident diabetes through PBPs. PBPs were calculated by deducting the pro- (primary outcome) for approximately 3 years after jected value of the Core (over the 5 years) from the ap- randomization (www.d2dstudy.org). Cancer and cardiovas- proved 5-year budget and dividing the balance by the cular disease are key secondary outcomes. The design of projected site enrollment to establish a site-specific PBP Lovegreen et al. Trials (2018) 19:267 Page 3 of 8 Table 1 Suggested breakdown by the D2d Coordinating Center Table 3 Scheme for Performance-Based Payments per participant of personnel effort for a typical study site that is projected to visit for a typical D2d site that is projected to enroll (i.e., randomize) enroll (i.e., randomize) and follow 125 participants and follow 125 participants Year 1 Year 2 Year 3 Year 4 Year 5 Per participant visit reimbursement (%) (%) (%) (%) (%) Visit % of total Performance-Based US$ Principal investigator 20 20 20 20 20 Payment amount and co-investigator(s) Month 00 20 982 (randomization) Research coordinator 100 100 100 75 75 Month 6 6 295 Research assistant 50 100 100 75 75 Assumptions: 125 participants will be randomized by year 2. The bulk of participant Month 12 20 982 follow-up is expected to occur in years 2–5 Month18 6 295 Month 24 20 982 amount per enrolled participant (Table 2). The total PBP Month 30 6 295 amount per enrolled participant was then distributed Month 36 22 1081 across seven scheduled study visits, proportionally to the Total 100 4912 work involved at each visit (Table 3). The Coordinating Month 42 6 295 Center issued quarterly PBPs to the sites based on actual visits conducted, verified by data in the study’selectronic Month 48 22 1081 data capture system. Based on this model, a site would only Total site-specific Performance-Based Payment amount per enrolled participant who completes all seven scheduled study visits, month 00 to month 36 (see also receive its total approved 5-year budget if the targeted text and Table 2). While the actual amount paid per participant visit varied by site number of participants were enrolled and followed for the (column 3), the percent of the total Performance-Based Payment amount allocated in each visit (column 2) was identical for all sites study duration. Visits beyond month 36 (month 42, etc.) are not included in the total site-specific Performance-Based Payment calculations becausetheaverageparticipant is expected to be followed for 36 months. A site that enrolls participants early in the recruitment period and follows these participants beyond month 36 will be Budget year 2: revisions to the model/addition of new reimbursed for additional visits (e.g., month 42 and month 48) in the same sites way as other similar visits (e.g., month 30 and month 36) The model was revised in year 2 to better accommodate the sites’ evolving needs. The central components (Core and PBPs) were preserved, but Core was redefined as 40% Budget year 3 and beyond: further revisions to the model of the total annual expenses (i.e., personnel effort and Because recruitment continued through year 3 and non-personnel costs) (Table 2). Quarterly PBPs continued part of year 4, the model was further revised. In years to be made to the sites based on successful enrollment of 3 and 4, sites received the same Core (in absolute US$) participants and completion of follow-up visits. as in year 2 (Table 2). The model reduced the planned Due to a delayed start of recruitment and lag in meeting Core in year 5, such that the total Core received over enrollment targets, D2d added four new sites in year 2. 5 years was the same as the original proposed Core The new sites submitted a standard 4-year NIH budget (Table 2). Sites were reminded that as the study pro- (grant years 2 to 5), and the revised methodology gressed they would receive a larger proportion of their described above was used to determine Core and PBPs. budget via PBPs. Table 2 Fundamental scheme of the hybrid financial management model utilized by D2d for a typical site that projected to enroll (i.e., randomize) 125 participants Year 1 Year 2 Year 3 Year 4 Year 5 Total Personnel (effort and fringe benefits) 178,135 183,100 183,100 137,171 137,171 Non-personnel 44,534 65,586 60,208 22,043 12,297 Total proposed budget 222,669 248,686 243,308 159,214 149,468 1,023,345 Core Proposed Core 89,068 91,550 91,550 68,586 68,586 409,339 Actual Core 89,068 99,474 99,474 99,474 21,848 409,339 Performance-Based Payments PBP amount per enrolled participant: 4912 614,007 Amounts shown in US$; numbers include directsand 53%indirectrate. Theschemeisflexibleboth from year to year and from site to site, while maintaining consistency Proposed Core was defined at the start of the study as 50% of annual personnel costs Actual Core varied from year to year (see text) while the total amount over 5 years did not exceed the total proposed Core (i.e., US$409,339) Lovegreen et al. Trials (2018) 19:267 Page 4 of 8 Transition to PBP-only for underperforming sites US$7.4M would have been needed to complete recruitment The model implemented a uniform approach to manage (Fig. 2). If D2d had issued cost reimbursement subaward underperforming sites. If the D2d leadership determined agreements and sites invoiced for their entire budget, 83% that there was potential for recruitment to improve at of the award (US$33M) would have been spent compared an underperforming site, the Coordinating Center con- to 65% of the award spent (US$26M) under the hybrid tinued to provide Core support for a limited “grace model used by D2d over the first 4 years of the trial (Fig. 2). period” of 3, 4, or 6 months. During that time, the Coordinating Center worked closely with the underper- forming site to establish new, realistic enrollment goals Budget year 1: determination of Core vs. PBP value and supported the site to achieve the revised goals. In year 1, individual site budgets varied considerably If a site did not meet its new target, site payments (range US$3324 to US$8406 per participant to recruit transitioned to PBP-only. The PBP-only alternative elim- and follow), largely due to varying salaries for research inated Core but increased the PBP amount per enrolled staff and fringe benefits (range US$38,027 to US$82,000 participant accordingly, so that the site was not penal- for research coordinators; US$25,094 to US$45,000 for ized for continuing to enroll and follow participants. research assistants; 15 to 41% for fringe benefits). These differences were attributed to regional differences (i.e., Primary outcome New York City vs. smaller cities in the Midwest), staff The primary outcome of interest in relation to the seniority, and expertise level. hybrid model used by the D2d study was total expenses In year 1, site activation and participant recruitment under this model compared to total expenses under a were slower than anticipated, which had a trickle-down traditional model. effect on the model. Sites invoiced Core for 50% of actual personnel expenses, but could not fully recoup Results the remaining 50% through PBPs due to slow recruit- Recruitment occurred from October 2013 through ment. At the end of year 1, study-wide recruitment was December 2016, and 2423 participants were randomized. at 12% of the target (vs. 50% projected) and 12% of the The D2d clinical sites were located throughout the USA total grant award was invested. Of the amount invested (Fig. 1) and included academic non-profit private med- in year 1, 54% went to the clinical sites (46% as Core ical centers, public medical centers, or Veterans and 8% as PBPs) and 46% went to other components Administration medical centers. (Coordinating Center, central laboratory, and drug The hybrid model used by D2d worked well to complete distribution). During year 1, two sites were discontin- recruitment and have funds remaining to follow partici- ued, one due to personnel turnover and one for signifi- pants. Under a cost reimbursement model, an additional cant underperformance. Fig. 1 Location of collaborating clinical sites in the D2d study Lovegreen et al. Trials (2018) 19:267 Page 5 of 8 Fig. 2 Comparison of the hybrid model used by D2d vs. a cost reimbursement subaward agreement for clinical sites. The cost reimbursement model includes the originally submitted budgets for sites (as cost reimbursement agreements) and actual expenses for non-site components (Coordinating Center, drug distribution, central lab) through the end of the recruitment phase without any budget reductions for underperformance. The hybrid model used by D2d includes actual spending (clinical sites and non-site components) through the end of the recruitment phase Budget year 2: revisions to the model/addition of new sites Budget year 3 and beyond: further revisions to the model Revisions to the model and addition of new sites in year As recruitment continued beyond year 2, into years 3 2 were critical to the completion of recruitment. Revi- and 4, the changes in Core payments were critical to sions to the Core gave sites the flexibility to allocate maintain recruitment activities. Increasing Core pay- Core support towards any study-related expense to meet ments in years 3 and 4 and reducing Core payments in site goals. At the end of recruitment, half of the sites year 5 allowed recruitment to be completed within the met or exceeded the 80% mark of their recruitment tar- original budget. get (Fig. 3). As in many large, multi-center trials, there was a wide range of recruitment success. Median enroll- Transition to PBP-only for underperforming sites ment at each site was 88 participants (range 29–318; 20 Transitioning underperforming sites to PBP-only agree- to 205% of the site target; Fig. 3). Three out of four new ments did not have an appreciable effect on recruitment sites (75%) met or exceeded their enrollment target, vs. or retention. Sites that transitioned to PBP-only agree- four out of 18 (22%) for the original sites. Of the D2d ments enrolled an average of 1.05 participant per month cohort, 20% was accounted for by the four new sites. before the transition, compared to 1.35 participant per Fig. 3 Enrollment in the D2d study. Enrollment is shown as percent of target, study-wide (dark blue bar) and for each clinical site (light blue bars). Horizontal line indicates the 80% mark of the enrollment target. Number of randomized participants is shown at the top of each bar. (√) indicates new sites added in year 2. (*) indicates sites that transitioned to a Performance-Based Payment-only arrangement Lovegreen et al. Trials (2018) 19:267 Page 6 of 8 month after the transition. Sites that continued to re- electronic health record (EHR)-based recruitment ap- ceive Core and PBP randomized 2.5 participants per proach, especially for a trial with laboratory-based inclusion month. Sites receiving PBP-only completed 92% of all criteria. To allocate resources based on effort and results, participant encounters vs. 95% for sites receiving Core the model established expectations in terms of recruitment and PBPs. and retention targets and timelines. Site recruitment was monitored weekly by the study leadership, and the model encouraged sites to abandon inefficient approaches. The Discussion bulk of the funds provided to the sites matched recruitment The hybrid financial management model used in the D2d and retention results, especially as the study progressed. study was designed to promote judicious investment of At the outset, recognizing that few sites would reach their precious resources in a way that does not compromise goal, the Coordinating Center estimated that the average recruitment and follow-up of participants. The fundamen- site would enroll about 80% of its projected target, while a tal principle behind the model is to allocate the total few sites would exceed their target. The prediction was cor- award earmarked for the sites proportional to their effort rect, as exactly half of the sites met or exceeded the 80% and results, while maintaining open-mindedness and flexi- mark at the end of the recruitment period. However, the bility to meet sites’ evolving needs. The methodology used range of recruitment success (20 to 205% of the stated goal) to determine the Core and PBPs based on each site’s was larger than anticipated. Therefore, flexibility to divert budget of what they expected their actual study costs over funds away from underperforming sites to match needs of 5 years was novel. In the process, we learned several well-performing sites was a key component of the model. lessons (Table 4)that we discuss below. Maintain open-mindedness and flexibility when Encourage multi-disciplinary teamwork addressing sites’ concerns In D2d, the study’s Principal Investigator, Coordinating Although the hybrid payment model resulted in a more Center, Research Administration at the prime grantee efficient investment of research funds in direct propor- institution, and the funding agency worked collabora- tion to results, many sites faced budget deficiencies due tively to design and employ a principles-based financial to underperformance and unanticipated issues, including management model to ensure appropriate stewardship increases in fringe rate, facilities and administrative rate, of the available funds. The model aimed for consistency and other expenses not accounted for in their original while recognizing sites’ unique circumstances. At least budget. Sites also did not fully appreciate that they in part, the hybrid model used by D2d worked well would receive their entire approved 5-year budget only if because people from different disciplines established a they fully met their enrollment and retention goals. For close and dynamic collaboration and decisions were example, if a site reached 80% of its enrollment target made in a cooperative fashion with the clinical sites. and successfully completed all follow-up visits, its PBPs would be reduced proportionally by 20%. Allocate resources based on results, not just effort Recognizing sites’ unique circumstance is important in Although effort and results often correlate, the association large, multi-center trials and decisions need to be made in is far from perfect, as effort may be exerted inefficiently or a consistent manner. The D2d leadership kept an open ineffectively. For example, bulk mailing of non-targeted let- mind throughout the recruitment phase—the most variable ters is an inefficient recruitment method compared to an phase of a clinical trial—and worked individually with sites, recognizing the uniqueness of the challenges faced by each, and strived to be flexible. For example, in each year during Table 4 The D2d hybrid budget management model: lessons the recruitment phase, the D2d leadership used a uniform learned approach to adjust the Core to better accommodate sites’ � Encourage multi-disciplinary teamwork that engages the Principal needs. For outlier sites, the D2d leadership re-established Investigator, Coordinating Center, Research Administration at the expectations and adjusted the budget as needed. When prime, subaward institutions, and the funding agency. faced with the potential of a site withdrawing from the � Allocate resources based on results, not just effort. study due to financial issues, the D2d leadership worked � Maintain open-mindedness and flexibility when addressing sites’ with the site to modify recruitment goals and its budget, concerns. while staying within the framework of the model. � Provide Core support and sufficient resources for study start-up. � Accept that sites will be discontinued and sites will be added. Provide Core support and sufficient resources for study � Project that recruitment will take longer and the outcome event rate start-up may be lower than predicted. A key concept of the hybrid model used by D2d is to pro- � Reinforce the model’s principles to site staff. vide Core funds for study start-up (e.g., IRB submission, Lovegreen et al. Trials (2018) 19:267 Page 7 of 8 study setup) and partial salary support (e.g., pre-screening [8], which may require study extension, thereby adding of participants, participation in committees). While a strain to the budget. The hybrid model used by D2d con- PBP-only model is conservative and advantageous to the served funds to ensure completion of recruitment while total study budget, under such model, sites may be reluc- conserving funds to continue following participants. If tant to invest in the initial effort, including exploring in- D2d had issued cost reimbursement agreements and sites novative recruitment methods. Study-wide, providing invoiced for their entire budget without any modifications Core funds may be a risky approach if many sites under- for underperformance, nearly the entire grant award perform. In D2d, providing Core for a proportion of an- would have been spent at the completion of recruitment. nual personnel effort only did not work. Redefining the Core as 40% of the annual total budget (including both Reinforce the model’s principles to site staff personnel and non-personnel) worked well and allowed A hybrid model can be new to investigators and research sites to overcome administrative concerns. The Core com- administration staff; therefore, it is important to reinforce ponent proved to be critical to the success of D2d as it the principles of the model throughout the study. When allowed sites to invest in the successful launch of the trial. executing the original subawards, the Coordinating Center provided a written document that described the model. As Accept that sites will be discontinued and sites will be the model was enhanced each year, the D2d leadership added maintained open communication with sites to optimize Despite extensive experience and initial enthusiasm, a few understanding and application of the model, especially sites will never get off the ground and many will underper- given turnover of site research administrative staff. form. There are many reasons that can prevent sites from reaching their full potential, including natural disasters, Limitations research staff relocation or retirement, illness affecting key The model worked for D2d, which is a focused and personnel, and shifting priorities at the site. It is important highly structured trial with laboratory-based eligibility for the leadership team to identify sites that have the criteria and outcomes. The model may be more difficult potential to improve with appropriate guidance and sites to adapt in a more complicated trial with multiple out- that need to be discontinued. The decision to discontinue comes. Other groups are exploring different methods to a site is exceedingly difficult because of financial, human minimize costs of clinical trials, such as use of mobile (personnel), academic (collegial), and emotional implica- technology or electronic medical records [9, 10]. Add- tions. There are ethical concerns in relation to the investi- itionally, the model was applied in a trial supported by gators’ duty to enrolled participants. However, timely the NIH, and results may not apply to non-NIH trials or separation is necessary as delay only utilizes more funds those outside of the USA. and makes the discontinuation even more problematic, es- pecially when many participants need to be administra- Conclusions tively withdrawn. Discontinuation of sites needs to follow The NIH-supported D2d clinical trial employed a hybrid a consistent process, and the decision needs to be unani- financial management model, comprised of two compo- mous by the study leadership. nents (Core and Performance-Based Payments) to man- It is certain that not all original sites will perform as age the study’s budget in a dynamic and creative way to projected; therefore, adding new sites should be antici- effectively invest finite resources. The development and pated and planned. Adding sites while the trial is employment of the model is the result of a cooperative ongoing is logistically challenging; however, the leader- effort between Research Administration at the prime ship, based on accumulated study-specific experience, is grantee institution, Principal Investigator, Coordinating better prepared to evaluate and select new sites whose Center, and the primary funding agency. Anticipating a strengths match the study’s needs. The addition of new delay in recruitment, the model continuously evaluated sites proved critical to the success of D2d. sites’ needs and re-allocated resources to meet the study enrollment goal and succeeded in managing the study’s Recruitment will take longer and the outcome event rate funds during the extended recruitment period in a way may be lower than projected that did not compromise the trial’s scientific integrity. It is rare for large trials to complete recruitment within Abbreviations the original timeline. In D2d, the recruitment period was D2d: Vitamin D and Type 2 Diabetes; NIDDK: National Institute of Diabetes projected to last 2 years; however, despite the sites’ best and Digestive and Kidney Diseases; NIH: National Institutes of Health; PBPs: Performance-Based Payments efforts, including adoption of a very successful electronic health record-based recruitment approach, enrollment Acknowledgements was slower than projected. Furthermore, the rate of occur- We thank the D2d investigators, staff, and trial participants for their rence of the primary outcome can be lower than projected outstanding dedication and commitment to the study. Lovegreen et al. Trials (2018) 19:267 Page 8 of 8 Funding 4. Reardon S, Tollefson J, Witze A, Ross E. US science agencies face deep cuts The planning phase of D2d was funded by the National Institute of Diabetes in Trump budget. 2017. http://www.nature.com/news/us-science-agencies- and Digestive and Kidney Diseases (NIDDK) through a multi-center clinical study face-deep-cuts-in-trump-budget-1.21652. Accessed 7 Apr 2017. implementation planning grant (U34) to Tufts Medical Center in Boston, MA 5. Larson GS, Carey C, Grarup J, Hudson F, Sachi K, Vjecha MJ, Gordin F, Group (U34DK091958; principal investigator A.G.P.). Planning was also supported in I. Lessons learned: infrastructure development and financial management part by the Intramural Research Program of the NIDDK. The conduct of D2d is for large, publicly funded, international trials. Clin Trials. 2016;13:127–36. primarily supported by the NIDDK and the Office of Dietary Supplements of the 6. Pittas AG, Dawson-Hughes B, Sheehan PR, Rosen CJ, Ware JH, Knowler National Institutes of Health through a multi-center clinical study cooperative WC, Staten MA, Group DdR. Rationale and design of the Vitamin D and agreement (U01DK098245; principal investigator A.G.P.) to Tufts Medical Center, Type 2 Diabetes (D2d) study: a diabetes prevention trial. Diabetes Care. where the D2d Coordinating Center is based. The U01 grant mechanism 2014;37:3227–34. establishes the NIDDK project scientist (M.A.S.) as a member of the D2d 7. American Diabetes Association. Standards of medical care in diabetes–2010. Research Group. The study also received secondary funding from the Diabetes Care. 2010;33(Suppl 1):S11–61. American Diabetes Association (1-14-D2d-01). Educational materials are 8. Lappe J, Watson P, Travers-Gustafson D, Recker R, Garland C, Gorham E, provided by the National Diabetes Education Program. No pharmaceutical Baggerly K, McDonnell SL. Effect of vitamin D and calcium supplementation manufacturers contributed to the planning, design, or conduct of D2d. on cancer incidence in older women: a randomized clinical trial. JAMA. Study pills were purchased from an independent nutritional supplement 2017;317:1234–43. manufacturing company that has no association with any members of the 9. Eapen ZJ, Lauer MS, Temple RJ. The imperative of overcoming barriers to D2d Research Group. The D2d investigators and the NIDDK project scientist the conduct of large, simple trials. JAMA. 2014;311:1397–8. were responsible for the design and conduct of the study; collection, 10. Clinical Trials Transformation Initiative. Mobile clinical trials. https://www.ctti- management, analysis, and interpretation of the data; preparation, review, clinicaltrials.org/programs/mobile-clinical-trials. and approval of the manuscript; and decision to submit the manuscript for publication. Availability of data and materials This study is registered on ClinicalTrials.gov (NCT01942694). Data are not publically available but are available from the corresponding author on reasonable request and in accord with NIH regulations. Authors’ contributions All authors contributed to the design of the D2d financial management model. OL, DR, MAS, and AGP contributed to the execution of the model. All authors read and approved the final manuscript. Ethics approval and consent to participate The study is approved and monitored by an independent data and safety monitoring board (appointed by the National Institute of Diabetes and Digestive and Kidney Diseases) and the institutional review board of each collaborating clinical research site. Written informed consent was obtained from each participant prior to study procedures. Competing interests The authors declare that they have no competing interests. Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Author details Research Administration, Department of Medicine, Tufts Medical Center, 800 Washington St, Boston, MA 02111, USA. Kelly Government Solutions for National Institute of Diabetes and Digestive and Kidney Diseases, 900 Rockville Pike, Bethesda, MD 20892, USA. Division of Endocrinology, Diabetes and Metabolism, Department of Medicine, Tufts Medical Center, 800 Washington St, Boston, MA 02111, USA. Received: 7 August 2017 Accepted: 10 April 2018 References 1. Battelle Technology Partnership Practice. Biopharmaceutical industry- sponsored clinical trials: impact on state economies. 2015. http://phrma-docs. phrma.org/sites/default/files/pdf/biopharmaceutical-industry-sponsored- clinical-trials-impact-on-state-economies.pdf. 2. Hudson KL, Lauer MS, Collins FS. Toward a new era of trust and transparency in clinical trials. JAMA. 2016;316:1353–4. 3. Johnson J. Brief history of NIH funding: fact sheet. Congressional Research Service. Prepared for Members and Committes of Congress. 2013. https:// www.everycrsreport.com/files/20131223_R43341_73c480e1172aef191e8cb81 fc7ecbf895ad830b4.pdf.

Journal

TrialsSpringer Journals

Published: May 3, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off