Limited Resource Trials

Well-conducted clinical research trials that address questions relevant in resource limited settings (RLS), like the study examined here, are of critical importance in the fight against AIDS.¹ Nevirapine (NVP) is a drug commonly used in RLS to prevent mother to child transmission (pMTCT). For many mothers, its use in pMTCT leads to later emergence of resistance to NVP, which is also standard of care for first line treatment in many RLS. Study A5207 sought to identify interventions to reduce NVP resistance. However, rapid implementation of such time-sensitive multi-site studies is challenging.^{2, 3, 5} Ensuring cross-site standardization and expeditiously implementing clinical research in RLS remains encumbered by inadequate infrastructure (e.g., need for appropriate space, training, and major equipment) and overburdened health systems.^{1, 2, 6, 8} Most often laboratory, pharmacy, clinical, and regulatory areas in RLS bring about start-up challenges.

Laboratories play a critical role in the conduct of therapeutic clinical trials. They ensure not only the continuity of research data, but also quality of care and safety of participants on study.⁶ The Centers for Disease Control and Prevention (CDC) released a report in 2004 indicating the importance of and need for laboratory capacity building to support HIV/AIDS care programs in RLS.⁶ Although the report centered on laboratory services that were needed for the President's Emergency Plan for AIDS Relief, in which half the sites in this study have participated, CDC's findings suggested that laboratory capacity building in RLS is fraught with significant challenges needing to be addressed.

Clinical research pharmacies play an equally important role in the research process. Research pharmacies are responsible for dispensing study drug, counseling research participants, conducting pill counts, and ensuring security and destruction of the study drug. When study product is dispensed outside the research area, careful attention to details around storage and administration is needed.

Research conducted in hospitals also poses challenges. Hospitals in RLS often lack appropriate areas for storing study medications that require refrigeration. Clinical and pharmacy areas are frequently set up with open air ventilation, making it a challenge to find secure storage that is temperature controlled. Dependable power sources are needed for equipment and for alarm systems necessary to monitor temperature and provide security.

To ensure the protection of human research participants and the quality of data collected for the study, an institutional review board (IRB) or ethics committee (EC) must approve the protocol document and the site-specific informed consent.⁷ The IRB should have appropriate representation from the medical community and from the community under study,⁷ which aids in the effort to ensure research is guided by ethical principles derived from the Common Rule.

The AIDS Clinical Trials Group (ACTG) Network Operations Center works with scientists around the globe to develop and implement protocols that address questions relevant to people living with HIV and AIDS. The network began its first large international trial in 2004, the same year that the study A5207, "3 ART Strategies to Reduce NVP-Resistant HIV after Intrapartum Single Dose NVP," was released to sites. The A5207 study was designed to evaluate the effectiveness of three different antiretroviral strategies toward prevention of NVP resistance in mothers who received single-dose NVP at the onset of active labor. Pregnant women 13 years and older were enrolled. The A5207 sample size was 420 mother/infant pairs, 70 in each of three arms. The A5207 study was funded through the Division of AIDS (DAIDS) at the National Institute of Allergy and Infectious Diseases (NIAID).

The International Program and Laboratory Coordination Groups at ACTG Operations Center closely supported the development of sites and their laboratories for preparedness of studies such as A5207. Despite high levels of support, the last site opened the A5207 study in March 2009.

Methods

The study was conducted using semi-structured interviews and survey data obtained during the course of site preparation to participate in A5207.

Focused discussions with site personnel were also conducted to further investigate the causes of study scale-up delays. The interviews consisted of questions about preparedness, such as potential solutions to in-country approvals, laboratory, pharmacy, and clinical preparedness challenges. Interviews elicited descriptive responses and study scale-up lessons learned. Follow-up meetings revealed sites' solutions to challenges. Follow-up surveys were conducted during the course of capacity building activities at sites. Responses came from site investigators, study coordinators, data managers, pharmacists, nurses, laboratory directors, and regulatory personnel.

With the exception of nurses and regulatory personnel, at least one of each of the aforementioned professions from each site was represented. Investigators and lead site personnel were encouraged to review all data collected and comment on the findings.

Purpose

Protocol A5207 was designed for implementation in locations where single-dose NVP was the standard of care for pMTCT of HIV/AIDS. The purpose of this study was to assess challenges and lessons learned among the eight sites that were selected to participate in A5207, which were located in Haiti, India, Malawi, South Africa, Tanzania, and Uganda. India and South Africa had two sites in each country.

The A5207 protocol team and international program staff at the ACTG Operations Center assessed how study conduct would fit into regular practices, processes, and work flow at sites. The nature of the A5207 study made it essential to assess activities that were not part of the site's usual practices. For example, it was important to determine study drug chain of custody as well as inventory maintenance and storage conditions at each site. This team identified adjustments needed to support protocol implementation and worked to modify forms and processes for the pharmacy, laboratory, clinical, and regulatory areas to meet the needs of the study.

Challenges and solutions

Although diverse with respect to geography, infrastructure, previous research experience, and HIV standard of care, all sites had common development and study implementation issues. In all cases, challenges to one or all four primary targets areas (laboratory, pharmacy, clinical, and regulatory) affected study scale-up and implementation. Sites addressed difficult staffing conditions, inadequate infrastructure, and evolving regulations around human research and local HIV standards of care to successfully open A5207, as is outlined in Table 1.

Table 1. Sites addressed difficult staffing conditions, inadequate infrastructure, and evolving regulations around human research and local HIV standards of care to successfully open A5207.

Laboratory practices

The US Congress established the Clinical Laboratory Improvement Amendments (CLIA) in 1988 to ensure accuracy and reliability of patient test results in the United States. Because laboratories participating in A5207 lacked access to CLIA certification, NIH supported development of these and other ACTG laboratories to ensure they provided accurate and reliable data for study-mandated laboratory evaluations.

Six of eight laboratories participating in A5207 had not conducted US-government-funded studies prior to their work with ACTG. These laboratories had to identify, hire and train staff, conduct validation and reference range studies, and develop back-up plans. These five laboratories required instrument validation before conducting A5207. They also had to show proficiency and accuracy of chemistry and hematology tests. Normal ranges for adult chemistry and hematology results were provided by instrument manufacturers. However, these reference ranges were based on studies conducted outside of the setting in which they would be used. Thus, laboratories were required to conduct studies to determine normal reference ranges for their local adult and infant populations before conducting A5207. Although laboratory validations and proficiency testing take valuable time to complete, the outcome is often much needed improvement in accuracy of results.

In addition to the need to learn about study-mandated tests and validation procedures, laboratories in RLS also battled with complex import and export logistics, irregularity of power and water supply, and lack of qualified laboratory technicians. Two laboratories had fewer challenges because they had been heavily supported by US-based universities for more than two years prior to their involvement in A5207. However, most laboratories had considerable scale-up challenges to study preparation. For those laboratories in more rural areas this was in part due to difficulties finding qualified personnel and the need for additional safeguards, such as backup systems for testing participants on study, given their more isolated setting.

According to site personnel, corruption in government ranks in one country was common place, sometimes leading to import and export delays and loss of study materials such as surveys and reagents, which are critical to laboratory development. For example, supplies sent to the site arrived in another city and were never recovered, which led to additional cost and start-up delays. This is not to suggest that all customs officials are corrupt. However, without careful preparation, shipments may not easily find their way to the final destination. Therefore, it is very important to know what services are provided by the courier, their ports of entry, and processes. Site personnel explained the various importation needs, which often change and make it difficult to standardize shipment processes.

Laboratories in Tanzania and Uganda found it particularly difficult to find skilled staff due to the small pool of qualified applicants. This was even more difficult in rural settings where opportunities for laboratory education are sparse. Several laboratory managers reported that employees were often hired and trained only to have them leave for higher paying positions elsewhere. To address laboratory staffing needs, two sites consistently hired staff from the United States, because they better understood the complexity of US requirements. Both sites faced staff changes two-to-three years later. Alternatively, several sites also opted to employ Chinese, European, and US laboratory volunteers. Again, they were available for a limited time, usually two years, at which time the site would need replacements. Regular staff turnover adds to a lack of continuity and institutional memory.

Though developing proficiency for new tests that meet NIH standards and finding ways to address staff turnover continue to plague some laboratories, all of them overcame multiple barriers. They installed generators and backup generators to address fluctuations in the power supply. At least three sites employed agents to work with customs officials and to be present at the port of entry when shipments arrived to ensure supplies made it to the site. All laboratories attained a level of proficiency and established adult reference ranges so that they were able to begin study A5207.

Local staff members holding technical positions gained needed experience with study-mandated tests and developed an understanding of validation procedures.

Pharmacy services

A5207 required that study product be dispensed and administered to participants in several locations (e.g., labor and delivery, recovery, postpartum areas, and outpatient clinics) and at unscheduled times (i.e., onset of active labor). This raised several questions about:

• Where would study drugs be stored to ensure the conditions specified by the manufacturer were met?

• Who would administer the study drug?

• Where would the drug be administered?

• Where would documentation be stored so that confidentiality was maintained?

Sites participating in A5207 generally had their postpartum areas in open rooms that accommodated many patients. Because medications were administered in open rooms, it was difficult to maintain research participant confidentiality. Developing relationships between pharmacy, clinic, and labor and delivery ward staff became crucial to ensuring participant safety and confidentiality.

In order to identify and address issues related to confidentiality, study medication chain of custody, and maintenance of proper conditions for the medication, the protocol team required sites to complete a study implementation plan. Sites had to provide: a list of A5207 study staff that would transport the mother's A5207 medications to the labor and delivery area; a sample prescription and label of the study medication; and copies of the documentation required for study medication transport. The team pharmacist reviewed the site implementation plans and worked with pharmacists to resolve issues. Once issues were resolved, sites were approved to participate in the study.

In some cases, site pharmacists had limited experience in managing study medications and maintaining documentation of the chain of custody for the study participants' medications. For most pharmacies in RLS, standard practice is to use small plastic or paper bags to dispense patient medications. For A5207, the integrity of some of the study medications could have been compromised if dispensed in plastic bags. To ensure proper dispensing tools, the sponsor provided bottles for dispensing study medications.

Finally, unique challenges arose at sites that had not routinely imported study medications. The team quickly learned that the rules of import and export in RLS could change unexpectedly, as was experienced with importation of laboratory supplies. Researchers carefully assessed pharmacy-specific issues, as drug supply chain and drug importation requirements vary from country to country. At least one site staff member from each site was fully trained and knowledgeable about the processes for obtaining study drug importation documents for that country. Moreover, communication with US-based sponsors became an "understood" priority. Pharmacies for A5207 were considered ready to conduct the study when staff was in place, and successful importation of study drug had been accomplished at least once. Upon putting mechanisms in place, subsequent shipments seemed easier. However, importation regulations did change in at least one country during the initiation of the study, creating yet another start-up delay for that site.

Clinical research practices

Conducting this perinatal HIV transmission study required a coordinated effort to ensure successful implementation, specifically involving HIV specialists, physicians, obstetricians, midwives, pediatricians, nurses, and outreach workers—research personnel would need to be available for labor ward deliveries at any hour. At half of the sites, HIV care and maternal healthcare facilities were co-located, making for easy coordination (Table 2). The other four hospitals where mothers would deliver were located outside of the established research setting, in some cases geographically separated by several kilometers. The physical distance between research sites, antenatal clinics, and labor and delivery wards added to logistical complexity at these sites and posed challenges to good clinical research practices. Summarized in Table 2 are the development needs that sites faced when they began start-up activities for A5207.

Table 2. At half of the sites, HIV care and maternal healthcare facilities were co-located, making for easy coordination

In addition to conducting research, most sites provided patient care to large patient volumes in overburdened clinical systems. Antenatal clinics and labor wards required infrastructure improvements. For example, new buildings, patient wards, and the addition of major equipment (e.g., air conditioners and refrigerators) were needed at some sites in order to accommodate study requirements.

Coordinated communication among the various levels of healthcare workers was essential and needed to be arranged prior to implementation of the study. Coordination required clearly identifying persons responsible for performing study-mandated evaluations, on a 24-hour, seven-day-per-week basis. Two sites found staff reluctant to travel to the hospital after dark. The solution at two sites was to provide "on call" staff with a safe means of transportation should they be called out after dark. Staff in labor wards would identify study participants, notify research staff, and at some sites they would also collect a baseline blood sample for pharmacokinetic analysis and administer the study drug.

However, at three sites, clinical staff turnover and motivation were problematic. Labor and delivery wards were only loosely affiliated with the CRSs. Three sites found hospital workers in an already busy labor and delivery ward did not appreciate the added burden of research responsibilities. Site leadership addressed this issue by educating staff about the study, providing incentives to keep labor and delivery staff engaged, and rendering active support from the research staff.

Sites also had to come up with innovative ways to address study procedures and the need to ensure drug storage temperature and security control in labor and delivery areas. Sites fashioned the needed space by building cabinets, closing in closets, and installing refrigerators and alarms where needed. They developed plans for facilitating transportation to the hospital for study mothers to ensure timely administration of study drug. Each site had also developed innovative ways to ensure that study participants presenting in labor were appropriately identified in a reliable but confidential manner. For example, at one site, intake staff was trained to identify women who presented using a simple sticker on their charts. Yet other sites had systems that instructed study participants to show their affiliation to the CRS with an identity card. These strategies helped to ensure staff communication, patient confidentiality, and documentation of study participant drop out due to administration of non-study pMTCT drug.

Regulatory affairs

The detailed documentation for the regulatory review process could test the patience of even the most steadfast researcher. When an ACTG protocol is released to sites, the investigator at each site is responsible for obtaining all required regulatory approvals. In settings where English is not the first language, translation and back-translation of the protocol informed consent document into the local language is required before IRB submissions. This process can take months. More than half the sites participating in this study used research staff to complete the translation and back-translation process of the informed consent. Subsequent to translation, the protocol and informed consent form are sent to local and national IRBs.

The IRB approval process is one of the biggest start-up challenges that sites in RLS conducting this study experienced. We found that institutional and national IRB review timelines were quite lengthy, taking an average of 37 months to complete (Table 3). The protracted timeline to obtaining regulatory approvals was rarely in the control of site administrators and investigators. Regulatory requirements delay sites working with US agencies. This is not unique to sites funded by US government agencies; European Union countries working in RLS also encounter similar delays.¹¹

Table 3. Timeline from Food and Drug Administration review to regulatory compliance center approval.

IRBs in RLS are often burdened with over-loaded staff, time constraints, and by the number of studies they are asked to review. Large workloads lead some IRBs to meeting irregularly and sometimes infrequently. For example, IRBs at two sites met every two or three months. Half of the sites were dealing with more than two regulatory entities. Six of the eight sites were required to send the protocol to their country's ministry of health (MOH). The sites in India required an additional approval because the research was being conducted by foreign investigators. Sites were also required to submit to drug regulatory bodies, functioning similarly to the US Food and Drug Administration. Three sites each had two collaborating US universities, and review was required from both. Changes required by one of the reviewing committees led to the need for approval from the other approving bodies. Therefore, in-country local, national, and US-based regulatory reviews required sequential submission.

While all other sites struggled to get through the regulatory maze of multiple IRBs, two of the three sites that did not have as many reviews were delayed when they had difficulty securing indemnity insurance for study participants and their fetuses, consuming months of time and energy in attempts to find coverage. The insurance that was finally obtained was very expensive.

The level of detail required for US regulatory submissions necessitates experience with US requirements and a meticulous attention to written detail. A tremendous amount of mentoring would be necessary to ensure quality regulatory submissions. As a result, more than half of the sites examined in this study used staff from the United States to manage regulatory affairs, introducing the same staff turnover challenges experienced with laboratory start-up, mentioned earlier. Despite an increase in familiarity with US regulations that has been obtained, the time it takes to get approval for new and amended protocols remains lengthy.

Conclusions

Given all of the logistical; regulatory; and ethical challenges, completion of the site start-up process for A5207, from release of the protocol to enrollment of the first participant, took an average of 33 months. Nonetheless, this initiative demonstrates the ability of sites in RLS to overcome tremendous challenges to succeed at building the infrastructure needed to implement new US government-funded HIV therapeutic studies. Requirements at sites in RLS seeking to conduct US-funded studies would have been even more difficult to meet without capacity building initiatives. Limited local resources for addressing those logistical issues raised in this case-study remain a demanding and challenging part of research implementation in RLS. Overarching capacity building themes include staff training and retention, regulatory complexity, and procurement difficulties.

While the financing of health workers is the single most costly and logistically difficult element of providing health services in general,^{8, 9, 10} staffing for research carries with it an additional responsibility. Research often demonstrates results that change policy and improve public health. Despite strides made toward building capacity at the sites in A5207, research staff identification, education, and retention remain significant issues. Finding and retaining a strong research workforce requires deliberate planning and competitive salaries.¹⁰ Leaders must engender creative solutions to staff retention problems by seriously considering salary increases, sharing human resources with patient care facilities, and providing incentives, such as training and transportation support to staff.

A trend toward requiring indemnity insurance in many African countries may also prove problematic, as was discovered in the course of study start-up for A5207. Indemnity plans were available on a limited basis and imposed unanticipated exorbitant costs on sites. Sponsors could potentially shorten start-up timelines by assisting with identification of affordable indemnity insurance plans early in the process for clinical trial participants. US sponsors who are unable to pay for indemnity insurance might expedite studies by negotiating with carriers to reduce premiums.

Realistic projections for study initiation should allow sufficient time for in-country IRB and MOH approvals and subsequent US-based approvals, if necessary. Long timelines for study start-up in RLS will continue to hamper US-sponsored research as ECs work towards harmonization of guidelines across multiple countries.

Missing from the discourse about timelines to regulatory approval are the cultural, political, and historical elements that influence decision-making. Resource deficiencies that approving institutions often face in RLS, could be influenced by historical perspectives. For example, attitudes toward Western society may affect decisions related to approving science conducted by US researchers. In some cases, IRBs may struggle with a history in their country of "parachute" researchers (scientists come to a country other than theirs, do a study, and leave without giving back to that country or community) that led them to carefully scrutinize and question the purpose of US-funded research. By engaging local investigators the ACTG has assisted non-US sites to build research capacity and to overcome some of these barriers, if in no other way than persistence.

Finally, those organizations intent on sponsoring clinical trials at sites in RLS with little or no experience responding to regulations outside their country should be prepared to ensure the site's capacity to participate before measuring its capabilities to perform. Sites in RLS are met with infrastructural and logistical challenges above and beyond those faced by their counterparts in non-RLS, affecting community, laboratory, pharmacy, clinical, and regulatory areas of clinical research. If evaluation involves comparing sites in RLS to clinics in non-RLS, evaluators should use rigorous scientific evaluation to ensure equal capacity to perform before making a comparison. Otherwise, sites in RLS may need weighted consideration during the evaluation process.

Although not covered here, it should be noted that all ACTG Network sites are required to have a functioning community advisory board (CAB). Community engagement ensures researchers understand local perspectives about feasibility and acceptability of studies. New CABs at sites involved in this study had start-up challenges such as the need for education and support.

With the increasing emphasis on HIV/AIDS therapeutic research in RLS, researchers should recognize the unique complexities involved in multi-site study implementation in these settings. Failure to address barriers to rapid start-up at sites in RLS has a cascade effect that can impact the timely dissemination of important study results, which could have serious implications for healthcare delivery and policy recommendations. There is a clear need for a systematic approach to address infrastructure deficiencies and delay of in-country regulatory review procedures, and as yet to be explored, the impact of HIV research capacity on healthcare delivery systems.

Christina Blanchard-Horan,* PhD, MA, is International Program Specialist, e-mail: cblanchardhoran@gmail.com, Vicki Stocker, RN, MS, is Section Manager, Protocol Management, and Laura Moran, MPH, is Clinical Trials Specialist all at Social & Scientific Systems, Inc., Silver Spring, MD. Elaine Okubo Ferguson, RPh, MS, is a Pharmacist at Food and Drug Administration, Bethesda, MD. Karin L. Klingman, MD, Medical Officer, National Institute of Allergy and Infectious Diseases (NIAID), Bethesda, MD. Deborah McMahon, MD, is Associate Professor of Medicine, University of Pittsburgh, Pittsburgh, PA. Jane Hitti, MD, MPH, is Associate Professor, Department of Obstetrics and Gynecology at University of Washington, Seattle, WA.

*To whom all correspondence should be addressed.

Editor's Note: The tables for this article are availabe online at appliedclinicaltrialsonline.com.

References

1. M. A. Lansang and R. Dennis, "Building Capacity in Health Research in the Developing World," Bulletin of the World Health Organization, 84 (10) 764–770 (2004).

2. A. Schulz-Baldes, E. Vayena, and N. Biller-Andorno, "Sharing Benefits in International Health Research: Research-Capacity Building as an Example of an Indirect Collective Benefit," European Molecular Biology Organization Reports, 8 (1) 8-13 (2007).

3. A. J. Bell, K. Brattegaard, M. Oziemkowska, T. K. Thomas, L. Slutsker, K. M. DeCock, J. Vulule, P. N. Amornkul, and A. Kombo, "Lessons Learned in Setting up a Clinical Trial in a Resource-Limited Area Without Previous Clinical Trials Experience," International Conference on AIDS, Atlanta, Georgia, July 11-16, 2004.

4. S. Crouse, "Ethics in Public Health Research," Public Policy and Ethics, 94 (6) 918-922 (2004).

5. K. Page-Shafer, V. Saphonn, L. P. Sun, M. C. Vun, D. A. Cooper, and J. M. Kalor, "HIV Prevention Research in Resource-Limited Setting: The Experience of Planning a Trial in Cambodia," The Lancet, 366 (9495) 1,499-1,503, (2005).

6. Centers for Disease Control and Prevention, "Building Laboratory Capacity in Support of HIV/AIDS Care Programs in Resource-Limited Countries," report from Global AIDS Program meeting, Atlanta, Georgia December 16-17, 2003, http://wwwn.cdc.gov/dls/ILA/cd/BuildingLabCapacity/documents/report.pdf.

7. US Department of Health and Human Services, "Federalwide Assurance (FWA) for the Protection of Human Subjects: Terms of the Federalwide Assurance for the Protection of Human Subjects," http://www.hhs.gov/ohrp/assurances/assurances/filasurt.html.

8. L. R. Hirschhorn, L. Oguda, A. Fullem, N. Dreesch, and P. Wilson, "Estimating Health Workforce Needs for Antiretroviral Therapy in RLS," Human Resources for Health, 4 (1) (2006).

9. World Health Organization, "Global Health Workforce Alliance, Strategic Plan," (2008), http://www.who.int/workforcealliance/GHWA_STRATEGIC%20PLAN_ENGLISH_WEB.pdf.

10. World Health Organization, "The World Health Report 2006— Working Together for Health," (2006), http://www.who.int/whr/2006/en.

11. P. Matsoso, M. Auton, S. Banoo Fomundam, H. Leng, S. Noazin, amd B. Meyer, "How Does the Regulatory Framework Affect Incentives for Research and Development?" http://www.who.int/intellectualproperty/studies/regulatory_framework/en/index.html.