Donegan, Elizabeth MD*; Moncada, Jeanne MT†; Chernesky, Max PhD‡; Schachter, Julius PhD†
OVER A 5-YEAR PERIOD, we have been involved in a number of studies in Indonesia, which included as a goal the improvement of sexually transmitted infection (STI)/HIV laboratory capability. When visiting laboratories that had received donations, it was common to find expired and/or improperly stored reagents and test kits. Laboratory technologists were performing tests by following instructions in package inserts written in a foreign language without benefit of quality control procedures. Duplicated, unused (even unpacked), broken, and discarded laboratory equipment was seen during site visits. Equipment could have been nonfunctional because of improper maintenance, or initial reagent supplies could have been exhausted without provision for a sustainable reagent supply.
These experiences, and discussions with others working on developing countries’ projects, have led us to write this commentary to discuss the issues of providing consistent and sustainable laboratory capacities for the diagnosis of STI and HIV in developing countries. Diagnostic laboratory capacity is limited but important in Indonesia and many other developing countries. Country Ministries of Health working with international agencies determine STI/HIV infection rates and oversee diagnosis and treatment. They design and evaluate targeted social interventions and allocate resources based, in part, on these results. The laboratory results are also needed for surveillance purposes and as programmatic measures of success. The international agencies have also attempted to strengthen in-country laboratory activities and capabilities by providing training and donating expensive laboratory reagents and equipment. In some cases, they have also included technology transfer with the hope of achieving a sustainable laboratory presence. This type of support is practical and is done worldwide, but, unfortunately, these donations often have minimal and temporary impact on laboratory capacity.
This growing process, which can consume large amounts of monies, needs to be examined as part of each initiative. We address 4 aspects of laboratory function that need to be examined: 1) laboratory management, 2) appropriate equipment selection and ongoing maintenance, 3) technical training and test result accuracy, and 4) end-user education and use of test results. There could be other use–effectiveness areas to discuss; however, we hope that focusing on these 4 areas will serve as a catalyst for further discussion within this journal or elsewhere.
Generally speaking, donor program administrators and their staff do not have clinical laboratory expertise, but they plan, administer, and supervise laboratory improvement programs. The programs approach capacity-building by laboratory renovations, equipment and supply purchases, and the use of short-term consultants together with brief fellowships for selected local leaders. The problems we have seen with this approach in Indonesia are not unique to that country. In China, this method resulted in poor organization and leadership, unsatisfactory utilization and maintenance of equipment, and inconsistent pediatric care. After identifying the problems and taking corrective actions, the China project appears to have been successful by providing long-term technical support over a 4-year period. 1
Long-term, in-country technical support for selected laboratory specialists (both clinical pathologists and senior technologists), who are then responsible for training their in-country colleagues, is fundamental to improving laboratory capacity. Technical consultants can expand the laboratory management skills of local laboratory specialists and serve as an ongoing resource after their departure. These consultants can also develop inter- and intracountry ties and help regional laboratory specialists develop formal education programs. Project funding goals can be achieved with their advice as to the economic use of available laboratory resources.
Additionally, a fee-for-service awareness needs to be developed if technology transfer is to result in a sustainable laboratory presence. Developing-country governmental laboratories are often unfamiliar with cost accounting and test pricing because laboratory equipment and reagents have been purchased centrally. Models for delivering laboratory results in a cost-effective way need to be studied and could be unique to the setting. The question of what happens when the study or the donated tests are finished should be addressed at the beginning of the program.
Capacity-building for a sustainable laboratory includes provision of techniques for cost accounting and negotiation with distributors and donors. These concepts can be understood and promoted when agencies purchase “test results” and offer laboratory management advice as an alternative to simply donating test kit reagents and equipment.
Appropriate Selection and Maintenance of Equipment
Although most developing-country diagnostic initiatives will focus on the provision of inexpensive and simple tests for the field and laboratory, the eventual outcome will usually lead to equipment purchases. Program plans that include potential equipment purchase need to initially determine the purpose, use, volume, and frequency of proposed tests. Alternative test methods should be investigated before equipment is selected. Often, automated, highly technical, expensive equipment has a limited effective lifespan. Test methodology can become obsolete as technology improves over time. Often, the following issues are not adequately addressed: the type of power source needed (110/60 Hz vs. 220/50 Hz), the type of plug connections required, the availability of affordable electrical power, the presence of back-up generators for power blackouts, and the need for equipment climate control with stable air conditioning or a processed water source. Equipment, in addition to being appropriate to the environment, must be in working order, be well maintained, kept supplied with appropriate reagents, and be used by properly trained technologists. 2,6–8 For example, in Indonesia, 1 donor-funded program purchased refurbished equipment from the United States as a cost-saving measure. On receipt, the shipment lacked 1 essential component, and several units were non-operational. The missing component and the spare parts were unavailable in the country and, thus, had to be reordered. As a consequence, the apparent cost savings were not realized, and there were long unanticipated delays. By the time the equipment was in working order, the reagents had expired and the term of the program had been completed. The majority of the units were never used.
If cost is the major and only factor taken into account in choosing between quotes from competing suppliers, it can be more expensive in the long run. If there are local distributors, established laboratories can recommend and advise about their performance. Reliability of supply, distribution, and delivery time, 2 maintenance of equipment and parts in the country must be taken into consideration. 2,3 In a cost–benefit analysis, a slightly higher quote could be better because of reliability and technical support that can be obtained. 2,3
An attractive alternative is the purchase of reagents that have an equipment rental agreement. Although expensive, it can guarantee the maintenance of working laboratory instruments. These contracts have the benefit that as the technology changes, the laboratory can update to state-of-the-art equipment. Negotiated costs per tests can be lowered if the laboratory is a long-term, large-volume user. Because reagent costs are often beyond the means of provincial government laboratories and the individuals who would benefit from diagnostic test results, donor agencies could initially negotiate lower cost per tests for future use. The eventual goal is the development of affordable and sustainable testing programs involving group purchase agreements to create large-volume discounts. In addition, the concept of “public health” pricing versus “private sector” pricing should be discussed with the distributors. 4
There should be provisions made for the identification and solution of equipment problems. Routine maintenance of equipment is necessary and cost-effective. 9 Typically, a commitment of 4% to 6% of the annual laboratory budget is needed to service, repair, and maintain equipment. Unfortunately, donor agencies seldom fund the recurrent costs of routine maintenance, repair, and replacement. Government funds for these activities are usually limited and well below the annual laboratory budget needed. 9 Most developing countries cannot deal with equipment quality control without external technical and financial assistance. 9 Recognizing these issues, donor agencies could pay for equipment repair, on a case-by-case basis, contributing to “donor dependency” without providing routine maintenance. Another more costly option is to include a “lifetime” maintenance and repair contract when donating equipment. 2,10 Alternatively, the World Health Organization has developed repair and maintenance manuals for laboratory technicians 11 and has supported in-country training programs. 3,12 A final option, potentially sustainable, is to support standardized equipment lists (facilitating repairs and maintenance), centralized acquisition of equipment, and a government-run central fund for maintenance and repair of laboratory equipment. 3,9
Technical Training and Accurate Test Performance
The planning process for laboratory programs includes provision for ongoing technical training, and development and monitoring of quality control and quality assurance programs. STI/HIV laboratories also need standard operation procedures for laboratory tests, universal precautions, and disposal of biohazard materials.
Sustainable testing should be a primary goal. When donor funds stop, testing often does not resume unless there is other funding. In this case, test procedures could require revalidation, reacquisition of quality control samples, and perhaps retraining of technologists. Erratic testing schedules could generate technical problems, like was found in an Indonesian screening program using nonamplified nucleic acid testing for Neisseria gonorrhoeae (GC) and Chlamydia trachomatis (CT) in female sex workers. After a 6-month lapse in testing, negative and positive control samples gave variable results. After several months, using both in- and out-of-country technical support, it became clear that the detection reagent sampling lines contained partially crystallized reagents.
Laboratory programs must teach and then assure adherence to standard operating procedures (SOP) and quality control (QC) guidelines. In 1 STI clinic, we discovered (after 1 year) that samples for nonamplified GC/CT testing were being collected and held over a 1-month period before transfer to the laboratory even though SOP for this test required testing within 7 days of collection. In another clinic, samples were collected in a remote location and stored improperly. Such practices, if not monitored and corrected, compromise validity of test results, and result in poor clinical care and inaccurate estimates of infection.
Funding agencies need to assure that the technologist who runs a test is well trained in proper test performance and procedures. Technologists must know how to recognize, identify, and solve testing irregularities. 5 Often, donor agencies fund off-site, out-of-country training for key country personnel as a program side benefit without targeting the individual who performs the test. These training sessions are attended by the laboratory directors or principal investigators and are conducted in an environment so dissimilar to the proposed test site as to be nontranslatable. On-site training should allow more accurate transfer of technology. Certainly, laboratory directors must be familiar with test procedures, and use and must troubleshoot problems. However, the critical training is that of the technologist. One donor-training program designed and funded a program to develop a series of STI clinics and simple laboratories within Indonesia. The program funded a 1-month, classroom-based, out-of-country STI diagnosis and treatment program for 2 general practitioners, 1 of whom had limited English language skills. A laboratory technologist who had worked in a clinical laboratory in a remote area, without experience or training in an STI laboratory, was hired as the laboratory expert. The program administrators then sent the 2 physicians and the laboratory technologists to train personnel working in program-funded clinics in diagnosis and treatment of STIs, as well as simple STI laboratory techniques. Direct smears stained with Giemsa stain and wet mounts with saline and potassium hydroxide were used to “enhance” syndromic management. When on-site interpretation of stained slides collected during the training program was reviewed in a provincial laboratory, 80% of direct smears had been misinterpreted. The samples collected for nonamplified GC/CT testing were improperly transported and stored and could not be tested. The accuracy of diagnosis and the appropriateness of treatment were compromised.
Quality control/assurance and test validation programs are uncommon in many developing-country laboratories. The importance of QC and test validation needs to be stressed. Quality control testing begins with test validation panels, followed by parallel testing, and continues with within-run unknown positive and negative samples and testing programs that periodically assay a battery of more difficult unknown samples submitted by outside agencies. Masked testing of known samples (validation panel) should be used for each test procedure before test start-up in each laboratory. Initially, parallel testing of 20% of test samples in an established, well-controlled laboratory brings to light procedural problems. In our experience, this practice has revealed unacceptably high false-negative test results for both syphilis and CT infections in 2 different Indonesian studies. Many of the standard QC laboratory procedures used in resource-rich environments are not routine in resource-limited countries, but could be addressed through the World Health Organization or nongovernmental organization programs. For most Indonesian government laboratories, outside QC monitoring is biannual at best with little or no within-run controls. A laboratory's QC program should include control of individual tests, equipment, and the environment. One Indonesian laboratory reported a 50% drop in the prevalence of gonorrhea with a treatment program and deemed the program a great success. In fact, the prevalence rate had not changed. A new CO2 incubator with a faulty CO2 detection board had been installed. The laboratory did not recognize that the new incubator was not receiving CO2 because daily chemical CO2 concentration measurements were not performed, and a CO2-sensitive gonococcal QC organism was not routinely cultured. After we investigated the potential causes for the rapid decrease in the number of isolates and consulted with the supplier, the CO2 board was replaced, an appropriate monitoring system was established, and the number of GC isolates increased to the previous number of isolates per month. Unfortunately, this process took several months.
Pipettes need regular calibration; the laboratory, refrigerators, and freezers need the temperature to be monitored and controlled; and the rotators and centrifuges need to be speed-checked. Expiration dates and storage requirements for reagents and test kits must be followed. An effective quality assurance program provides a wider view of the impact of a test result on patient or end-user and extends outside of the laboratory. A quality assurance program examining the relevant, timely report and use of test results also is needed. Each laboratory needs to be monitored by experienced technical personnel specifically trained in these procedures until these practices become routine. Donor agencies, at times, mistakenly assume that performance and test monitoring is appropriate if performed at all, and they assume that passing 1 panel will guarantee future reliability. Ongoing QC programs are mandatory.
End-User Education and Use of Test Results
Availability of high-quality laboratory results does not assure appropriate test requests, and accurate test results reported quickly do not assure they will be interpreted correctly or accurately related to the patient in a timely manner. Appropriate treatment (if available) might not be initiated, even if a positive test result is available. Patient follow up needs to be monitored by increasing laboratory capacity to provide quality assurance activities. Treating physicians and those under their direction need to be educated as to the meaning and interpretation of both negative and positive laboratory results. This is particularly the case when using screening HIV antibody tests for diagnostic purposes. In Indonesia, reporting nonconfirmed false-positive screening HIV test results as positive has been documented (Dr. July Kumalawati Widyaharsana, DMM, personal communication, September 2001).
Lack of public confidence in positive test results from laboratories performing screening tests later found to be negative by more specific or confirmatory tests has also occurred (Dr. July Kumalawati Widyaharsana, DMM, personal communication). Additionally, screening programs for GC/CT and syphilis in Indonesia frequently do not locate and appropriately treat infected individuals. Despite widespread concern with infertility in Indonesia, few physicians ever test for either GC or CT because many physicians believe these infections can be diagnosed clinically using syndromic management algorithms.
Many donor-funded STI/HIV programs do not have full-time staff with laboratory backgrounds. It is likely that program development, selection and purchase of equipment, and negotiation with suppliers and laboratories will be the responsibility of a staff member without a background in either clinical engineering or laboratory work. “In most developing countries, new equipment is usually ordered by people who have no technical training with the result that some of the equipment will be unsuitable.”13 Unfortunately, the literature that could help to guide a public health medical officer through the issues is generally located in technical laboratory journals and books or in clinical engineering journals. Medical journals normally read by STI/HIV program staff have often provided little or no practical guidance, which could explain, in part, why well-documented lessons have not been incorporated into existing programs. We think there are roles to be played within funding and advocacy programs, and greater priority should be given to locating experienced technical advice on these issues. 2 These lessons learned in Indonesia might seem obvious, but they are also too often overlooked. We welcome comments from the STD Journal readership.