Pooling sera to reduce the cost of HIV surveillance: a feasibility study in a rural Kenyan district

Pooling as a strategy for the timely diagnosis of soil-transmitted helminths in stool: value and reproducibility

BACKGROUND

The strategy of pooling stool specimens has been extensively used in the field of parasitology in order to facilitate the screening of large numbers of samples whilst minimizing the prohibitive cost of single sample analysis. The aim of this study was to develop a standardized reproducible pooling protocol for stool samples, validated between two different laboratories, without jeopardizing the sensitivity of the quantitative polymerase chain reaction (qPCR) assays employed for the detection of soil-transmitted helminths (STHs). Two distinct experimental phases were recruited. First, the sensitivity and specificity of the established protocol was assessed by real-time PCR for each one of the STHs. Secondly, agreement and reproducibility of the protocol between the two different laboratories were tested. The need for multiple stool sampling to avoid false negative results was also assessed. Finally, a cost exercise was conducted which included labour cost in low- and high-wage settings, consumable cost, prevalence of a single STH species, and a simple distribution pattern of the positive samples in pools to estimate time and money savings suggested by the strategy.

RESULTS

The sensitivity of the pooling method was variable among the STH species but consistent between the two laboratories. Estimates of specificity indicate a 'pooling approach' can yield a low frequency of 'missed' infections. There were no significant differences regarding the execution of the protocol and the subsequent STH detection between the two laboratories, which suggests in most cases the protocol is reproducible by adequately trained staff. Finally, given the high degree of agreement, there appears to be little or no need for multiple sampling of either individuals or pools.

CONCLUSIONS

Our results suggest that the pooling protocol developed herein is a robust and efficient strategy for the detection of STHs in 'pools-of-five'. There is notable complexity of the pool preparation to ensure even distribution of helminth DNA throughout. Therefore, at a given setting, cost of labour among other logistical and epidemiological factors, is the more concerning and determining factor when choosing pooling strategies, rather than losing sensitivity and/or specificity of the molecular assay or the method.

Determination of Varying Group Sizes for Pooling Procedure

Pooling is an attractive strategy in screening infected specimens, especially for rare diseases. An essential step of performing the pooled test is to determine the group size. Sometimes, equal group size is not appropriate due to population heterogeneity. In this case, varying group sizes are preferred and could be determined while individual information is available. In this study, we propose a sequential procedure to determine varying group sizes through fully utilizing available information. This procedure is data driven. Simulations show that it has good performance in estimating parameters.

Regression analysis for multiple-disease group testing data

Group testing, where individual specimens are composited into groups to test for the presence of a disease (or other binary characteristic), is a procedure commonly used to reduce the costs of screening a large number of individuals. Group testing data are unique in that only group responses may be available, but inferences are needed at the individual level. A further methodological challenge arises when individuals are tested in groups for multiple diseases simultaneously, because unobserved individual disease statuses are likely correlated. In this paper, we propose new regression techniques for multiple-disease group testing data. We develop an expectation-solution based algorithm that provides consistent parameter estimates and natural large-sample inference procedures. We apply our proposed methodology to chlamydia and gonorrhea screening data collected in Nebraska as part of the Infertility Prevention Project and to prenatal infectious disease screening data from Kenya.

Comparison of individual and pooled stool samples for the assessment of soil-transmitted helminth infection intensity and drug efficacy

Background

In veterinary parasitology samples are often pooled for a rapid assessment of infection intensity and drug efficacy. Currently, studies evaluating this strategy in large-scale drug administration programs to control human soil-transmitted helminths (STHs; Ascaris lumbricoides, Trichuris trichiura, and hookworm), are absent. Therefore, we developed and evaluated a pooling strategy to assess intensity of STH infections and drug efficacy.

Methods/Principal Findings

Stool samples from 840 children attending 14 primary schools in Jimma, Ethiopia were pooled (pool sizes of 10, 20, and 60) to evaluate the infection intensity of STHs. In addition, the efficacy of a single dose of mebendazole (500 mg) in terms of fecal egg count reduction (FECR; synonym of egg reduction rate) was evaluated in 600 children from two of these schools. Individual and pooled samples were examined with the McMaster egg counting method. For each of the three STHs, we found a significant positive correlation between mean fecal egg counts (FECs) of individual stool samples and FEC of pooled stool samples, ranging from 0.62 to 0.98. Only for A. lumbricoides was any significant difference in mean FEC of the individual and pooled samples found. For this STH species, pools of 60 samples resulted in significantly higher FECs. FECR for the different number of samples pooled was comparable in all pool sizes, except for hookworm. For this parasite, pools of 10 and 60 samples provided significantly higher FECR results.

Conclusion/Significance

This study highlights that pooling stool samples holds promise as a strategy for rapidly assessing infection intensity and efficacy of administered drugs in programs to control human STHs. However, further research is required to determine when and how pooling of stool samples can be cost-effectively applied along a control program, and to verify whether this approach is also applicable to other NTDs.

Since the last decade, growing awareness of the control of neglected tropical diseases (NTDs) has resulted in worldwide increased pledges of drug donations. However, health care decision makers have a limited repertoire of strategies for a rapid assessment of infection intensity and for checking of drug resistance development. Therefore, we verified whether examination of pooled stool samples provide estimates of intestinal worm infection intensity and drug efficacy comparable to those obtained by examination of individual stool samples. Overall, the results showed that pooled samples provide comparable levels of infection intensity and drug efficacy. We conclude that pooling stool samples holds promise as a means of rapidly appraising the intensity of intestinal worm infections on a population level and of monitoring the efficacy of donated drugs. However, this study was conducted in an endemic region. Further research is required to determine when and how pooling of stool samples can be cost-effectively applied in a control program that is reducing the transmission of disease, and to verify whether this approach is also applicable to NTDs other than studied in this paper.

Sample size under inverse negative binomial group testing for accuracy in parameter estimation

BACKGROUND

The group testing method has been proposed for the detection and estimation of genetically modified plants (adventitious presence of unwanted transgenic plants, AP). For binary response variables (presence or absence), group testing is efficient when the prevalence is low, so that estimation, detection, and sample size methods have been developed under the binomial model. However, when the event is rare (low prevalence <0.1), and testing occurs sequentially, inverse (negative) binomial pooled sampling may be preferred.

METHODOLOGY/PRINCIPAL FINDINGS

This research proposes three sample size procedures (two computational and one analytic) for estimating prevalence using group testing under inverse (negative) binomial sampling. These methods provide the required number of positive pools ([Formula: see text]), given a pool size (k), for estimating the proportion of AP plants using the Dorfman model and inverse (negative) binomial sampling. We give real and simulated examples to show how to apply these methods and the proposed sample-size formula. The Monte Carlo method was used to study the coverage and level of assurance achieved by the proposed sample sizes. An R program to create other scenarios is given in Appendix S2.

CONCLUSIONS

The three methods ensure precision in the estimated proportion of AP because they guarantee that the width (W) of the confidence interval (CI) will be equal to, or narrower than, the desired width ([Formula: see text]), with a probability of [Formula: see text]. With the Monte Carlo study we found that the computational Wald procedure (method 2) produces the more precise sample size (with coverage and assurance levels very close to nominal values) and that the samples size based on the Clopper-Pearson CI (method 1) is conservative (overestimates the sample size); the analytic Wald sample size method we developed (method 3) sometimes underestimated the optimum number of pools.

Global goodness-of-fit tests for group testing regression models

In a variety of biomedical applications, particularly those involving screening for infectious diseases, testing individuals (e.g. blood/urine samples, etc.) in pools has become a standard method of data collection. This experimental design, known as group testing (or pooled testing), can provide a large reduction in testing costs and can offer nearly the same precision as individual testing. To account for covariate information on individual subjects, regression models for group testing data have been proposed recently. However, there are currently no tools available to check the adequacy of these models. In this paper, we present various global goodness-of-fit tests for regression models with group testing data. We use simulation to examine the small-sample size and power properties of the tests for different pool composition strategies. We illustrate our methods using two infectious disease data sets, one from an HIV study in Kenya and one from the Infertility Prevention Project.

Bias, efficiency, and agreement for group-testing regression models

Group testing involves pooling individual items together and testing them simultaneously for a rare binary trait. Whether the goal is to estimate the prevalence of the trait or to identify those individuals that possess it, group testing can provide substantial benefits when compared to testing subjects individually. Recently, group-testing regression models have been proposed as a way to incorporate covariates when estimating trait prevalence. In this paper, we examine these models by comparing fits obtained from individual and group testing samples. Relative bias and efficiency measures are used to assess the accuracy and precision of the resulting estimates using different grouping strategies. We also investigate the agreement of individual and group-testing regression estimates for various grouping strategies and the effects of group size selection. Depending on how groups are formed, our results show that group-testing regression models can perform very well when compared to the analogous models based on individual observations. However, different grouping strategies can provide very different results in finite samples.

Evaluation of the dried blood spot filter paper technology and five testing strategies of HIV-1 and HIV-2 infections in West Africa

Simple robust approaches are needed to monitor the prevalence and incidence of HIV in Africa. The aim of this study was to evaluate the use of dried blood spot (DBS) as an alternative to serum or plasma for sentinel surveillance. Paired DBS and blood samples were obtained from 200 patients attending a genito-urinary medicine clinic in West Africa. The gold standard of diagnosis was based on the combination of 3 enzyme-linked immunosorbent assays (ELISA) using serum. The presence of HIV antibodies in eluates of dried blood spots was detected by ELISA, Gelatin Particle Assay (GPA) and Pepti-Lav 1-2 in 5 different testing strategies. All 200 eluates were tested individually, and in addition pools of 5 eluates each were tested. The sensitivity of the testing strategies ranged from 95.0% (83.1 - 99.4%) to 100% and the specificity from 97.5% (93.7 - 99.3%) to 100%. Testing in pools of 5 did not affect sensitivity. Dried blood spots were easy to work with. Test kit and laboratory consumable costs varied between 492 pounds and 1037 pounds (unpooled strategies) and 163 pounds and 421 pounds (pooled). The monospecific ELISAs used in this study are no longer in production; currently available differentiating assays need to be tested. DBS are recommended for sentinel surveillance in Africa.

Hypothesis tests for and against a simple order among proportions estimated by pooled testing

The use of pooled testing as a means of estimating the prevalence of rare traits has received considerable attention in recent years, particularly in the areas of public health, genetics, animal-disease assessment, and plant pathology. In pooled-testing applications, observations are made on pools of individuals amalgamated together. In this paper, we examine order-restricted hypothesis tests involving k >2 binomial proportions estimated by pooled testing, extending the earlier work of Tebbs and Swallow (2003, Biometrika, 90, 471-477 and Biometrical Journal, 45, 618-630). In particular, we focus on (i) testing the equality of proportions versus an isotonic alternative and (ii) testing for a violation of isotonicity. We propose new tests for each scenario and provide results which characterize the small-sample performance of our procedures. We illustrate our methods using two data sets; one from an observational HIV study and one from an agricultural experiment.

Accuracy of hepatitis C virus core antigen testing in pools among seroconverters

BACKGROUND

Screening blood units for hepatitis C virus (HCV) with nucleic acid testing (NAT) reduces the risk associated with the long "window period" (8-9 weeks) after HCV infection. The feasibility of adding the HCV core antigen assay in pools to the existing anti-HCV individual screening was examined as an alternative of NAT, for early detection of HCV.

STUDY DESIGN AND METHODS

Eighteen HCV seroconversion panels were tested for HCV antibodies, HCV antigen, and HCV RNA. Each sample was tested for HCV antigen individually and in pools of 3, 6, and 12. Statistical analyses included estimation of time until detection of the first positive HCV antigen bleed in each pool size, with a locally weighted regression (LOWESS) model. Sensitivity was calculated compared to NAT.

RESULTS

Detection of HCV antigen in individual samples and in pools of 3 and 6 significantly preceded the detection of antibodies by 63, 53, and 46 days, respectively. Although the sensitivity of the HCV antigen test decreased with the increase in pool size, the estimated overall sensitivity of the "two-stage" antigen and antibody screening (where NAT of individual samples was the gold standard) was not significantly different between individual and the different pool sizes.

CONCLUSION

Screening for HCV antigen in pools of 6 can be considered an efficient and easier-to-implement alternative to the costly NAT for identifying blood donors in the seroconversion period. It may offer a cost-effective approach in resource utilization in poor countries, that, after the implementation of HCV antibody testing, want to further improve blood safety.

Pooling Experiments for Blood Screening and Drug Discovery

Pooling experiments date as far back as 1915 and were initially used in dilution studies for estimating the density of organisms in some medium. These early uses of pooling were necessitated by scientific and technical limitations. Today, pooling experiments are driven by the potential cost savings and precision gains that can result, and they are making a substantial impact on blood screening and drug discovery. A general review of pooling experiments is given here, with additional details and discussion of issues and methods for two important application areas, namely, blood testing and drug discovery. The blood testing application is very old, from 1943, yet is still used today, especially for HIV antibody screening. In contrast, the drug discovery application is relatively new, with early uses occurring in the period from the late 1980s to early 1990s. Statistical methods for this latter application are still actively being investigated and developed through both the pharmaceutical industries and academic research. The ability of pooling to investigate synergism offers exciting prospects for the discovery of combination therapies.

The effect of pooling sera on the detection of avian pneumovirus antibodies using an enzyme-linked immunosorbent assay test

Pooling of samples is a cost-effective approach to estimate disease prevalence and to identify infected individuals. The objective of this study was to evaluate the use of serum pools for the detection of avian pneumovirus infection in turkey flocks by enzyme-linked immunosorbent assay, so that a minimum number of tests can be performed without compromising the sensitivity and specificity of the test. A total of 900 field samples were tested; 20 samples from each of 45 flocks. All samples were tested individually followed by pool testing in groups of 3, 4, 5, and 7 samples each. The number of positive pools for a given pool size was positively associated with the number of positive samples. In a separate experiment, the effect of dilution was examined by pooling 1 positive sample with different numbers of negative samples to form pools of sizes 2-7. These laboratory results were analyzed and integrated into a simulation model aimed at evaluating cost-efficient testing procedures. The probability of detecting an infected flock depended on prevalence of infection, size of serum pool, and the cutoff value used for optical density difference. At a theoretical prevalence of 20%, the probability of detecting an infected flock was 0.93 and 0.86 for a pool of 2 and 7, respectively. The probability of detecting positive flocks increased with increased prevalence and decreased cutoff. Pooling of samples represented a significant reduction in the cost of testing, suggesting that pooling is more advantageous and cost effective than testing individual samples.

Pooling of Chlamydia laboratory tests to determine the prevalence of ocular Chlamydia trachomatis infection

With the Global Elimination of Trachoma by 2020 program underway, it has become increasingly important to identify the prevalence of ocular chlamydia infection in communities. DNA amplification tests are the gold standard, but are prohibitively expensive. In the present paper, we investigate whether pooling multiple specimens into a single test is feasible. The conjunctivae of 170 children in western Nepal were examined and swabbed. The prevalence of chlamydial infection was estimated in two ways using the ligase chain reaction: by testing all 170 specimens individually, and by testing 34 pools of 5 specimens each. We show that the confidence interval for 34 pooled specimens approaches that of doing all 170 specimens as the prevalence decreases. We also determine the optimal number of specimens to pool into a single test to minimize the confidence interval of the estimate. If the population prevalence is expected to be around 10%, then 14 specimens should be pooled per test. Even at 50% prevalence, costs can be reduced by pooling two samples per test.

Regression models for disease prevalence with diagnostic tests on pools of serum samples

Whether the aim is to diagnose individuals or estimate prevalence, many epidemiological studies have demonstrated the successful use of tests on pooled sera. These tests detect whether at least one sample in the pool is positive. Although originally designed to reduce diagnostic costs, testing pools also lowers false positive and negative rates in low prevalence settings and yields more precise prevalence estimates. Current methods are aimed at estimating the average population risk from diagnostic tests on pools. In this article, we extend the original class of risk estimators to adjust for covariates recorded on individual pool members. Maximum likelihood theory provides a flexible estimation method that handles different covariate values in the pool, different pool sizes, and errors in test results. In special cases, software for generalized linear models can be used. Pool design has a strong impact on precision and cost efficiency, with covariate-homogeneous pools carrying the largest amount of information. We perform joint pool and sample size calculations using information from individual contributors to the pool and show that a good design can severely reduce cost and yet increase precision. The methods are illustrated using data from a Kenyan surveillance study of HIV. Compared to individual testing, age-homogeneous, optimal-sized pools of average size seven reduce cost to 44% of the original price with virtually no loss in precision.