Adjusting mortality for loss to follow-up: analysis of five ART programmes in sub-Saharan Africa

Background

Evaluation of antiretroviral treatment (ART) programmes in sub-Saharan Africa is difficult because many patients are lost to follow-up. Outcomes in these patients are generally unknown but studies tracing patients have shown mortality to be high. We adjusted programme-level mortality in the first year of antiretroviral treatment (ART) for excess mortality in patients lost to follow-up.

Methods and Findings

Treatment-naïve patients starting combination ART in five programmes in Côte d''Ivoire, Kenya, Malawi and South Africa were eligible. Patients whose last visit was at least nine months before the closure of the database were considered lost to follow-up. We filled missing survival times in these patients by multiple imputation, using estimates of mortality from studies that traced patients lost to follow-up. Data were analyzed using Weibull models, adjusting for age, sex, ART regimen, CD4 cell count, clinical stage and treatment programme. A total of 15,915 HIV-infected patients (median CD4 cell count 110 cells/µL, median age 35 years, 68% female) were included; 1,001 (6.3%) were known to have died and 1,285 (14.3%) were lost to follow-up in the first year of ART. Crude estimates of mortality at one year ranged from 5.7% (95% CI 4.9–6.5%) to 10.9% (9.6–12.4%) across the five programmes. Estimated mortality hazard ratios comparing patients lost to follow-up with those remaining in care ranged from 6 to 23. Adjusted estimates based on these hazard ratios ranged from 10.2% (8.9–11.6%) to 16.9% (15.0–19.1%), with relative increases in mortality ranging from 27% to 73% across programmes.

Conclusions

Naïve survival analysis ignoring excess mortality in patients lost to follow-up may greatly underestimate overall mortality, and bias ART programme evaluations. Adjusted mortality estimates can be obtained based on excess mortality rates in patients lost to follow-up.     

The frequency of endemic fetal loss in dairy cattle: a review

Estimates of the frequency of endemic fetal loss range from 0.4 to 10.6%, a 26-fold difference, in 26 studies of dairy herds over 5 decades. Sources of this difference include breed, geographic, study population, case definition and procedural differences. The definition of fetal loss was inconsistent among the studies and was often not clearly stated. Twelve of the studies cited only visually observed losses, while the remainder reported losses of all known pregnancies but using different gestational at-risk periods. Definitions of the resulting frequency measure and methods of computation varied widely among the studies. Fetal loss frequencies have been reported as ratios, proportions, incidence rates and cumulative incidence rates. Denominators used in frequency calculations have included the number of pregnant cows, number of cows in the breeding herd, number of cows that calved before or after the risk period and number of calves born. These procedural differences make comparisons among studies and an overall estimate of an expected frequency of endemic fetal loss in dairy cattle difficult at best. Procedural and definition differences aside, the median frequency of fetal loss from studies enumerating only observed abortions was 1.95% and from the studies that included both observed and unobserved abortions 6.5%. Based on this difference, approximately only 30% of endemic fetal losses are observed visually.     

Correcting mortality for loss to follow-up: a nomogram applied to antiretroviral treatment programmes in sub-Saharan Africa

Background

The World Health Organization estimates that in sub-Saharan Africa about 4 million HIV-infected patients had started antiretroviral therapy (ART) by the end of 2008. Loss of patients to follow-up and care is an important problem for treatment programmes in this region. As mortality is high in these patients compared to patients remaining in care, ART programmes with high rates of loss to follow-up may substantially underestimate mortality of all patients starting ART.

Methods and Findings

We developed a nomogram to correct mortality estimates for loss to follow-up, based on the fact that mortality of all patients starting ART in a treatment programme is a weighted average of mortality among patients lost to follow-up and patients remaining in care. The nomogram gives a correction factor based on the percentage of patients lost to follow-up at a given point in time, and the estimated ratio of mortality between patients lost and not lost to follow-up. The mortality observed among patients retained in care is then multiplied by the correction factor to obtain an estimate of programme-level mortality that takes all deaths into account. A web calculator directly calculates the corrected, programme-level mortality with 95% confidence intervals (CIs). We applied the method to 11 ART programmes in sub-Saharan Africa. Patients retained in care had a mortality at 1 year of 1.4% to 12.0%; loss to follow-up ranged from 2.8% to 28.7%; and the correction factor from 1.2 to 8.0. The absolute difference between uncorrected and corrected mortality at 1 year ranged from 1.6% to 9.8%, and was above 5% in four programmes. The largest difference in mortality was in a programme with 28.7% of patients lost to follow-up at 1 year.

Conclusions

The amount of bias in mortality estimates can be large in ART programmes with substantial loss to follow-up. Programmes should routinely report mortality among patients retained in care and the proportion of patients lost. A simple nomogram can then be used to estimate mortality among all patients who started ART, for a range of plausible mortality rates among patients lost to follow-up. Please see later in the article for the Editors'' Summary     

Modelling the Spread of Farming in the Bantu-Speaking Regions of Africa: An Archaeology-Based Phylogeography

We use archaeological data and spatial methods to reconstruct the dispersal of farming into areas of sub-Saharan Africa now occupied by Bantu language speakers, and introduce a new large-scale radiocarbon database and a new suite of spatial modelling techniques. We also introduce a method of estimating phylogeographic relationships from archaeologically-modelled dispersal maps, with results produced in a format that enables comparison with linguistic and genetic phylogenies. Several hypotheses are explored. The ‘deep split’ hypothesis suggests that an early-branching eastern Bantu stream spread around the northern boundary of the equatorial rainforest, but recent linguistic and genetic work tends not to support this. An alternative riverine/littoral hypothesis suggests that rivers and coastlines facilitated the migration of the first farmers/horticulturalists, with some extending this to include rivers through the rainforest as conduits to East Africa. More recently, research has shown that a grassland corridor opened through the rainforest at around 3000–2500 BP, and the possible effect of this on migrating populations is also explored. Our results indicate that rivers and coasts were important dispersal corridors, but do not resolve the debate about a ‘Deep Split’. Future work should focus on improving the size, quality and geographical coverage of the archaeological ¹⁴C database; on augmenting the information base to establish descent relationships between archaeological sites and regions based on shared material cultural traits; and on refining the associated physical geographical reconstructions of changing land cover.     

Universal definition of loss to follow-up in HIV treatment programs: a statistical analysis of 111 facilities in Africa, Asia, and Latin America

Background

Although patient attrition is recognized as a threat to the long-term success of antiretroviral therapy programs worldwide, there is no universal definition for classifying patients as lost to follow-up (LTFU). We analyzed data from health facilities across Africa, Asia, and Latin America to empirically determine a standard LTFU definition.

Methods and Findings

At a set “status classification” date, patients were categorized as either “active” or “LTFU” according to different intervals from time of last clinic encounter. For each threshold, we looked forward 365 d to assess the performance and accuracy of this initial classification. The best-performing definition for LTFU had the lowest proportion of patients misclassified as active or LTFU. Observational data from 111 health facilities—representing 180,718 patients from 19 countries—were included in this study. In the primary analysis, for which data from all facilities were pooled, an interval of 180 d (95% confidence interval [CI]: 173–181 d) since last patient encounter resulted in the fewest misclassifications (7.7%, 95% CI: 7.6%–7.8%). A secondary analysis that gave equal weight to cohorts and to regions generated a similar result (175 d); however, an alternate approach that used inverse weighting for cohorts based on variance and equal weighting for regions produced a slightly lower summary measure (150 d). When examined at the facility level, the best-performing definition varied from 58 to 383 d (mean = 150 d), but when a standard definition of 180 d was applied to each facility, only slight increases in misclassification (mean = 1.2%, 95% CI: 1.0%–1.5%) were observed. Using this definition, the proportion of patients classified as LTFU by facility ranged from 3.1% to 45.1% (mean = 19.9%, 95% CI: 19.1%–21.7%).

Conclusions

Based on this evaluation, we recommend the adoption of ≥180 d since the last clinic visit as a standard LTFU definition. Such standardization is an important step to understanding the reasons that underlie patient attrition and establishing more reliable and comparable program evaluation worldwide. Please see later in the article for the Editors'' Summary     

Towards the selection of cattle for tick resistance in Africa

Half-body tick collections and visual assessment of tick burdens were performed monthly over six months on 100 bulls at the Kenya National Boran Stud, Mutara Ranch, Kenya.Boophilus decoloratus predominated among several tick species infesting the animals. Burdens ofB. decoloratus and total tick burdens were highly correlated. Rankings of relative tick resistance among bulls were consistent from month to month. Rankings based on visual assessment were very close to those based on actual tick burdens. Animals with thin skins appeared to carry fewer ticks, but tick burden bore no relationship to coat colour. The results suggest that simple visual inspection of total tick burdens may be a suitable basis for the selection of tick resistance in cattle.     

Spread of Q fever within dairy cattle herds: key parameters inferred using a Bayesian approach

Q fever is a worldwide zoonosis caused by Coxiella burnetii. Although ruminants are recognized as the most important source of human infection, no previous studies have focused on assessing the characteristics of the bacterial spread within a cattle herd and no epidemic model has been proposed in this context. We assess the key epidemiological parameters from field data in a Bayesian framework that takes into account the available knowledge, missing data and the uncertainty of the observation process owing to the imperfection of diagnostic tests. We propose an original individual-based Markovian model in discrete time describing the evolution of the infection for each animal. Markov chain Monte Carlo methodology is used to estimate parameters of interest from data consisting of individual health states of 217 cows of five chronically infected dairy herds sampled every week for a four-week period. Outputs are the posterior distributions of the probabilities of transition between health states and of the environmental bacterial load. Our findings show that some herds are characterized by a very low infection risk while others have a mild infection risk and a non-negligible intermittent shedding probability. Moreover, the antibody status seems to be a key point in the bacterial spread (shedders with antibodies shed for a longer period of time than shedders without antibodies). In addition to the biological insights, these estimates also provide information for calibrating simulation models to assess control strategies for C. burnetii infection.     

Evolution of culture-dependent discriminate sociality: a gene-culture coevolutionary model

Animals behave cooperatively towards certain conspecifics while being indifferent or even hostile to others. The distinction is made primarily according to kinship as predicted by the kin selection theory. With regards to humans, however, this is not always the case; in particular, humans sometimes exhibit a discriminate sociality on the basis of culturally transmitted traits, such as personal ornaments, languages, rituals, etc. This paper explores the possibility that the human faculty of cultural transmission and resultant cultural variation among individuals may have facilitated the evolution of discriminate sociality in humans. To this end, a gene-culture coevolutionary model is developed focusing on competition over control of resource as a context in which discriminate sociality may have evolved. Specifically, two types of culture-dependent discriminate sociality are considered: ingroup favouritism, with ingroup and outgroup being distinguished by the presence or absence of a cultural trait; and prestige hierarchies, with the prestige being conferred on the bearer of a cultural trait. The model specifies the conditions under which emergence and evolutionary stability of the two types of discriminate sociality are promoted by the presence of cultural variation among individuals.     

The role of cattle in the volatile loss of nitrogen from a shortgrass steppe

The cycling and volatile loss of N derived from cattle urine at upland and lowland sites within the shortgrass steppe of eastern Colorado was studied, using¹⁵N-labelled urea as an N source. Losses of NH ₀ ³ were determined by direct measurement and by difference. Losses were higher from coarse (27% summer, 12% winter) than from fine textured (0–2%) soils. Immobilization and plant uptake of N accounted for significant amounts of added N. Extrapolating our plot measurements to a typical pasture, using spatially and temporally stratified urine deposition data, losses from upland sites were calculated to be 0.016 g N · m^-2 · y^-1, while losses from lowland sites were negligible. This resulted in an average loss of 0.011 g N · m^-2 · y^-1 for a pasture divided 70:30 between uplands and lowlands. The loss of urine N calculated assuming no spatial stratification would be sevenfold higher (0.076 g N · m^-2 · y^-1). Losses of NH ₀ ³ from urine, animal biomass removal, and NH₂O loss totaled only 0.07 g N · m^-2 · y^-1 , or about 25% of wet deposition input. We calculated a potential loss of NH ₀ ³ from senescing vegetation of 0.26 g N · m^-2 · y^-1, an order of magnitude larger than all other losses combined.     

Counting the founders: the matrilineal genetic ancestry of the Jewish Diaspora

The history of the Jewish Diaspora dates back to the Assyrian and Babylonian conquests in the Levant, followed by complex demographic and migratory trajectories over the ensuing millennia which pose a serious challenge to unraveling population genetic patterns. Here we ask whether phylogenetic analysis, based on highly resolved mitochondrial DNA (mtDNA) phylogenies can discern among maternal ancestries of the Diaspora. Accordingly, 1,142 samples from 14 different non-Ashkenazi Jewish communities were analyzed. A list of complete mtDNA sequences was established for all variants present at high frequency in the communities studied, along with high-resolution genotyping of all samples. Unlike the previously reported pattern observed among Ashkenazi Jews, the numerically major portion of the non-Ashkenazi Jews, currently estimated at 5 million people and comprised of the Moroccan, Iraqi, Iranian and Iberian Exile Jewish communities showed no evidence for a narrow founder effect, which did however characterize the smaller and more remote Belmonte, Indian and the two Caucasus communities. The Indian and Ethiopian Jewish sample sets suggested local female introgression, while mtDNAs in all other communities studied belong to a well-characterized West Eurasian pool of maternal lineages. Absence of sub-Saharan African mtDNA lineages among the North African Jewish communities suggests negligible or low level of admixture with females of the host populations among whom the African haplogroup (Hg) L0-L3 sub-clades variants are common. In contrast, the North African and Iberian Exile Jewish communities show influence of putative Iberian admixture as documented by mtDNA Hg HV0 variants. These findings highlight striking differences in the demographic history of the widespread Jewish Diaspora.     

The matrilineal ancestry of Ashkenazi Jewry: portrait of a recent founder event

Both the extent and location of the maternal ancestral deme from which the Ashkenazi Jewry arose remain obscure. Here, using complete sequences of the maternally inherited mitochondrial DNA (mtDNA), we show that close to one-half of Ashkenazi Jews, estimated at 8,000,000 people, can be traced back to only 4 women carrying distinct mtDNAs that are virtually absent in other populations, with the important exception of low frequencies among non-Ashkenazi Jews. We conclude that four founding mtDNAs, likely of Near Eastern ancestry, underwent major expansion(s) in Europe within the past millennium.     

Coevolution: the geographic mosaic of coevolutionary arms races

Coevolutionary arms races between species can favor exaggeration of traits for attack and defense, but relentless escalation of these arms races does not necessarily occur in all populations.     

Spread of extensively drug-resistant tuberculosis in KwaZulu-Natal province, South Africa

Background

In 2005 a cluster of 53 HIV-infected patients with extensively drug-resistant tuberculosis (XDR-TB) was detected in the Msinga sub-district, the catchment area for the Church of Scotland Hospital (CoSH) in Tugela Ferry, in KwaZulu-Natal province (KZN), South Africa. KZN is divided into 11 healthcare districts. We sought to determine the distribution of XDR TB cases in the province in relation to population density.

Methods

In this cross-sectional study, the KZN tuberculosis laboratory database was analysed. Results of all patients with a sputum culture positive for Mycobacterium tuberculosis from January 2006 to June 2007 were included. Drug-susceptibility test results for isoniazid, rifampicin, ethambutol, streptomycin, kanamycin and ofloxacin were available for all patients as well as the location of the hospital where their clinical diagnosis was made.

Findings

In total, 20858 patients attending one of 73 hospitals or their adjacent clinics had cultures positive for M. tuberculosis. Of these, 4170 (20%) were MDR-TB cases. Four hundred and forty three (11%) of the MDR tuberculosis cases displayed the XDR tuberculosis susceptibility profile. Only 1429 (34%) of the MDR-TB patients were seen at the provincial referral hospital for treatment. The proportion of XDR-TB amongst culture-confirmed cases was highest in the Msinga sub-district (19.6%), followed by the remaining part of the Umzinyati district (5.9%) and the other 10 districts (1.1%). The number of hospitals with at least one XDR-TB case increased from 18 (25%) to 58 (80%) during the study period.

Interpretation

XDR-TB is present throughout KZN. More than 65% of all diagnosed MDR-TB cases, including XDR-TB patients, were left untreated and likely remained in the community as a source of infection.     

Diagnosis of cattle diseases endemic to sub-Saharan Africa: evaluating a low cost decision support tool in use by veterinary personnel

Background

Diagnosis is key to control and prevention of livestock diseases. In areas of sub-Saharan Africa where private practitioners rarely replace Government veterinary services reduced in effectiveness by structural adjustment programmes, those who remain lack resources for diagnosis and might benefit from decision support.

Methodology/Principal Findings

We evaluated whether a low-cost diagnostic decision support tool would lead to changes in clinical diagnostic practice by fifteen veterinary and animal health officers undertaking primary animal healthcare in Uganda. The eight diseases covered by the tool included 98% of all bovine diagnoses made before or after its introduction. It may therefore inform proportional morbidity in the area; breed, age and geographic location effects were consistent with current epidemiological understanding. Trypanosomosis, theileriosis, anaplasmosis, and parasitic gastroenteritis were the most common conditions among 713 bovine clinical cases diagnosed prior to introduction of the tool. Thereafter, in 747 bovine clinical cases estimated proportional morbidity of fasciolosis doubled, while theileriosis and parasitic gastroenteritis were diagnosed less commonly and the average number of clinical signs increased from 3.5 to 4.9 per case, with 28% of cases reporting six or more signs compared to 3% beforehand. Anaemia/pallor, weakness and staring coat contributed most to this increase, approximately doubling in number and were recorded in over half of all cases. Finally, although lack of a gold standard hindered objective assessment of whether the tool improved the reliability of diagnosis, informative concordance and misclassification matrices yielded useful insights into its role in the diagnostic process.

Conclusions/Significance

The diagnostic decision support tool covered the majority of diagnoses made before or after its introduction, leading to a significant increase in the number of clinical signs recorded, suggesting this as a key beneficial consequence of its use. It may also inform approximate proportional morbidity and represent a useful epidemiological tool in poorly resourced areas.     

Breed relationships and definition in British cattle: a genetic analysis

The genetic diversity of eight British cattle breeds was quantified in this study. In all, 30 microsatellites from the FAO panel of markers were used to characterise the DNA samples from nearly 400 individuals. A variety of methods were applied to analyse the data in order to look at diversity within and between breeds. The relationships between breeds were not highly resolved and breed clusters were not associated with geographical distribution. Analyses also defined the cohesiveness or definition of the various breeds, with Highland, Guernsey and Jersey as the best defined and most distinctive of the breeds.     

Molecular genetic analysis of a cattle population to reconstitute the extinct Algarvia breed

Background

Decisions to initiate conservation programmes need to account for extant variability, diversity loss and cultural and economic aspects. Molecular markers were used to investigate if putative Algarvia animals could be identified for use as progenitors in a breeding programme to recover this nearly extinct breed.

Methods

46 individuals phenotypically representative of Algarvia cattle were genotyped for 27 microsatellite loci and compared with 11 Portuguese autochthonous and three imported breeds. Genetic distances and factorial correspondence analyses (FCA) were performed to investigate the relationship among Algarvia and related breeds. Assignment tests were done to identify representative individuals of the breed. Y chromosome and mtDNA analyses were used to further characterize Algarvia animals. Gene- and allelic-based conservation analyses were used to determine breed contributions to overall genetic diversity.

Results

Genetic distance and FCA results confirmed the close relationship between Algarvia and southern Portuguese breeds. Assignment tests without breed information classified 17 Algarvia animals in this cluster with a high probability (q > 0.95). With breed information, 30 cows and three bulls were identified (q > 0.95) that could be used to reconstitute the Algarvia breed. Molecular and morphological results were concordant. These animals showed intermediate levels of genetic diversity (MNA = 6.0 ± 1.6, R_t = 5.7 ± 1.4, H_o = 0.63 ± 0.19 and H_e = 0.69 ± 0.10) relative to other Portuguese breeds. Evidence of inbreeding was also detected (F_is = 0.083, P < 0.001). The four Algarvia bulls had Y-haplotypes H6Y2 and H11Y2, common in Portuguese cattle. The mtDNA composition showed prevalence of T3 matrilines and presence of the African-derived T1a haplogroup. This analysis confirmed the genetic proximity of Algarvia and Garvonesa breeds (F_st = 0.028, P > 0.05). Algarvia cattle provide an intermediate contribution (CB = 6.18, CW = -0.06 and D1 = 0.50) to the overall gene diversity of Portuguese cattle. Algarvia and seven other autochthonous breeds made no contribution to the overall allelic diversity.

Conclusions

Molecular analyses complemented previous morphological findings to identify 33 animals that can be considered remnants of the Algarvia breed. Results of genetic diversity and conservation analyses provide objective information to establish a management program to reconstitute the Algarvia breed.