Quality of data collection in a large HIV observational clinic database in sub-Saharan Africa: implications for clinical research and audit of care

Lessons learned from over a decade of data audits in international observational HIV cohorts in Latin America and East Africa

Introduction

Routine patient care data are increasingly used for biomedical research, but such "secondary use" data have known limitations, including their quality. When leveraging routine care data for observational research, developing audit protocols that can maximize informational return and minimize costs is paramount.

Methods

For more than a decade, the Latin America and East Africa regions of the International epidemiology Databases to Evaluate AIDS (IeDEA) consortium have been auditing the observational data drawn from participating human immunodeficiency virus clinics. Since our earliest audits, where external auditors used paper forms to record audit findings from paper medical records, we have streamlined our protocols to obtain more efficient and informative audits that keep up with advancing technology while reducing travel obligations and associated costs.

Results

We present five key lessons learned from conducting data audits of secondary-use data from resource-limited settings for more than 10 years and share eight recommendations for other consortia looking to implement data quality initiatives.

Conclusion

After completing multiple audit cycles in both the Latin America and East Africa regions of the IeDEA consortium, we have established a rich reference for data quality in our cohorts, as well as large, audited analytical datasets that can be used to answer important clinical questions with confidence. By sharing our audit processes and how they have been adapted over time, we hope that others can develop protocols informed by our lessons learned from more than a decade of experience in these large, diverse cohorts.

Early evaluation of the transition from an analog to an electronic surgical logbook system in Sierra Leone

BACKGROUND

Surgical logbooks are a commonly used tool for quality assurance of surgical training. Electronic logbooks are increasingly applied in low-resource settings, but there is limited research on their quality. The aim of this study is to evaluate the quality of an app-based surgical e-logbook system shortly after its implementation in a low-income country and to identify potential areas of improvement for the system.

METHODS

Entries in the e-logbook system were cross-checked with hospital records and categorized as matched or overreported. Moreover, the hospital records were checked for underreported procedures. Additionally, semi-structured interviews were conducted with users of the e-logbook system.

RESULTS

A total of 278 e-logbook database entries and 379 procedures in the hospital records from 14 users were analyzed. Matches were found in the hospital records for 67.3% of the database entries. Moreover, 32.7% of the database entries were overreported and 50.7% of the procedures in the hospital records were underreported. A previous study of an analog surgical logbook system in the same setting estimated that 73.1% of the entries were matches or close matches. Interviews with 12 e-logbook users found overall satisfaction but also identified potential areas of improvement, including the need for more training in the use of the system, modifications to improve user-friendliness, and better access to the necessary technology.

CONCLUSIONS

A reliable documentation system is necessary to evaluate the quality of health workforce training. The early evaluation of a surgical e-logbook system in a low-income country showed that the collected data should be approached with caution. The quantitative analysis suggests that the e-logbook system needs to be improved in terms of accuracy. In interviews, users reported that digitalization of the logbook system was a much-needed innovation but also identified important areas of improvement. Recognition of these aspects at an early stage facilitates guidance and adjustment of further implementation and might improve the accuracy of the system.

Pregnancies among women living with HIV using contraceptives and antiretroviral therapy in western Kenya: a retrospective, cohort study

BACKGROUND

Preventing unintended pregnancies is paramount for women living with HIV (WLHIV). Previous studies have suggested that efavirenz-containing antiretroviral therapy (ART) reduces contraceptive effectiveness of implants, but there are uncertainties regarding the quality of the electronic medical record (EMR) data used in these prior studies.

METHODS

We conducted a retrospective, cohort study of EMR data from 2011 to 2015 among WLHIV of reproductive age accessing HIV care in public facilities in western Kenya. We validated a large subsample of records with manual chart review and telephone interviews. We estimated adjusted incidence rate ratios (aIRRs) with Poisson regression accounting for the validation sampling using inverse probability weighting and generalized raking.

RESULTS

A total of 85,324 women contributed a total of 170,845 women-years (w-y) of observation time; a subset of 5080 women had their charts reviewed, and 1285 underwent interviews. Among implant users, the aIRR of pregnancy for efavirenz- vs. nevirapine-containing ART was 1.9 (95% CI 1.6, 2.4) using EMR data only and 3.2 (95% CI 1.8, 5.7) when additionally using both chart review and interview validated data. Among efavirenz users, the aIRR of pregnancy for depomedroxyprogesterone acetate (DMPA) vs. implant use was 1.8 (95% CI 1.5, 2.1) in EMR only and 2.4 (95% CI 1.0, 6.1) using validated data.

CONCLUSION

Pregnancy rates are higher when contraceptive implants are concomitantly used with efavirenz-containing ART, though rates were similar to leading alternative contraceptive methods such as DMPA. Our data provides policymakers, program staff, and WLHIV greater confidence in guiding their decision-making around contraceptive and ART options. Our novel, 3-phase validation sampling provides an innovative tool for using routine EMR data to improve the robustness of data quality.

Improved generalized raking estimators to address dependent covariate and failure-time outcome error

Biomedical studies that use electronic health records (EHR) data for inference are often subject to bias due to measurement error. The measurement error present in EHR data is typically complex, consisting of errors of unknown functional form in covariates and the outcome, which can be dependent. To address the bias resulting from such errors, generalized raking has recently been proposed as a robust method that yields consistent estimates without the need to model the error structure. We provide rationale for why these previously proposed raking estimators can be expected to be inefficient in failure-time outcome settings involving misclassification of the event indicator. We propose raking estimators that utilize multiple imputation, to impute either the target variables or auxiliary variables, to improve the efficiency. We also consider outcome-dependent sampling designs and investigate their impact on the efficiency of the raking estimators, either with or without multiple imputation. We present an extensive numerical study to examine the performance of the proposed estimators across various measurement error settings. We then apply the proposed methods to our motivating setting, in which we seek to analyze HIV outcomes in an observational cohort with EHR data from the Vanderbilt Comprehensive Care Clinic.

Co-infection of Pneumocystis jirovecii pneumonia and pulmonary CMV in a infant with X-linked severe combined immunodeficiency

We herein report a rare case of co-infection of Pneumocystis jirovecii pneumonia and pulmonary CMV in a 3-month-old infant with X-linked severe combined immunodeficiency, in which diagnostic clues were obtained from the bronchoalveolar lavage fluid. We focus on the value of cytological diagnosis of P. jirovecii pneumonia and pulmonary CMV in the bronchoalveolar lavage fluid. Recognizing morphological characteristics of these pathogenic microorganisms is important to get timely diagnosis and treatment for the patients. Furthermore, repeated severe infections in infants should remind us to screen for immunosuppressed states.

Efficient semiparametric inference for two-phase studies with outcome and covariate measurement errors

In modern observational studies using electronic health records or other routinely collected data, both the outcome and covariates of interest can be error-prone and their errors often correlated. A cost-effective solution is the two-phase design, under which the error-prone outcome and covariates are observed for all subjects during the first phase and that information is used to select a validation subsample for accurate measurements of these variables in the second phase. Previous research on two-phase measurement error problems largely focused on scenarios where there are errors in covariates only or the validation sample is a simple random sample of study subjects. Herein, we propose a semiparametric approach to general two-phase measurement error problems with a quantitative outcome, allowing for correlated errors in the outcome and covariates and arbitrary second-phase selection. We devise a computationally efficient and numerically stable expectation-maximization algorithm to maximize the nonparametric likelihood function. The resulting estimators possess desired statistical properties. We demonstrate the superiority of the proposed methods over existing approaches through extensive simulation studies, and we illustrate their use in an observational HIV study.

Outcome Measurement at a Ugandan Referral Hospital: Validation of the Mbarara Surgical Services Quality Assurance Database

BACKGROUND

Five billion people lack access to surgery. Accurate and complete data have been identified as essential to the global scale-up of perioperative care. This study retrospectively validates the Mbarara Surgical Services Quality Assurance Database (SQUAD), an electronic outcomes database at a Ugandan secondary referral hospital.

METHODS

SQUAD data were compared to paper records from August 2013 to January 2017. To assess data entry accuracy, two researchers independently extracted 24 patient variables from 170 charts. To assess completeness of patient capture, SQUAD entries were compared to a sample of charts returned to the Medical Records Department, and to a sample of entries in ward and operating room logbooks. Two-tailed binomial proportions with 95% CI were calculated from the comparative results of patient observations, against a predefined accuracy of 0.85-0.95.

RESULTS

Agreement between completed validation observations from charts and SQUAD data was 91.5% (n = 3734/4080 data points). Binomial tests indicated that 15 variables had higher than 95% accuracy. A total 19 of 24 variables had ≥ 85% accuracy. The completeness of SQUAD patient capture was 98.2% (n = 167/170) of charts returned to the Medical Records Department, 97.5% (n = 198/203) of operating logbook entries, and 100% (n = 111/111) of ward logbook entries, respectively.

CONCLUSION

SQUAD closely reflects the primary surgical and anaesthetic data at a Ugandan secondary hospital. Data accuracy of key variables and completeness of population capture were comparable to those of databases in high-income countries and outperformed those of other low- and middle-income countries.

Errors in multiple variables in human immunodeficiency virus (HIV) cohort and electronic health record data: statistical challenges and opportunities

Objectives: Observational data derived from patient electronic health records (EHR) data are increasingly used for human immunodeficiency virus/acquired immunodeficiency syndrome (HIV/AIDS) research. There are challenges to using these data, in particular with regards to data quality; some are recognized, some unrecognized, and some recognized but ignored. There are great opportunities for the statistical community to improve inference by incorporating validation subsampling into analyses of EHR data.Methods: Methods to address measurement error, misclassification, and missing data are relevant, as are sampling designs such as two-phase sampling. However, many of the existing statistical methods for measurement error, for example, only address relatively simple settings, whereas the errors seen in these datasets span multiple variables (both predictors and outcomes), are correlated, and even affect who is included in the study.Results/Conclusion: We will discuss some preliminary methods in this area with a particular focus on time-to-event outcomes and outline areas of future research.

Self-audits as alternatives to travel-audits for improving data quality in the Caribbean, Central and South America network for HIV epidemiology

INTRODUCTION

Audits play a critical role in maintaining the integrity of observational cohort data. While previous work has validated the audit process, sending trained auditors to sites ("travel-audits") can be costly. We investigate the efficacy of training sites to conduct "self-audits."

METHODS

In 2017, eight research groups in the Caribbean, Central, and South America network for HIV Epidemiology each audited a subset of their patient records randomly selected by the data coordinating center at Vanderbilt. Designated investigators at each site compared abstracted research data to the original clinical source documents and captured audit findings electronically. Additionally, two Vanderbilt investigators performed on-site travel-audits at three randomly selected sites (one adult and two pediatric) in late summer 2017.

RESULTS

Self- and travel-auditors, respectively, reported that 93% and 92% of 8919 data entries, captured across 28 unique clinical variables on 65 patients, were entered correctly. Across all entries, 8409 (94%) received the same assessment from self- and travel-auditors (7988 correct and 421 incorrect). Of 421 entries mutually assessed as "incorrect," 304 (82%) were corrected by both self- and travel-auditors and 250 of these (72%) received the same corrections. Reason for changing antiretroviral therapy (ART) regimen, ART end date, viral load value, CD4%, and HIV diagnosis date had the most mismatched corrections.

CONCLUSIONS

With similar overall error rates, findings suggest that data audits conducted by trained local investigators could provide an alternative to on-site audits by external auditors to ensure continued data quality. However, discrepancies observed between corrections illustrate challenges in determining correct values even with audits.

The impact of data quality and source data verification on epidemiologic inference: a practical application using HIV observational data

BACKGROUND

Data audits are often evaluated soon after completion, even though the identification of systematic issues may lead to additional data quality improvements in the future. In this study, we assess the impact of the entire data audit process on subsequent statistical analyses.

METHODS

We conducted on-site audits of datasets from nine international HIV care sites. Error rates were quantified for key demographic and clinical variables among a subset of records randomly selected for auditing. Based on audit results, some sites were tasked with targeted validation of high-error-rate variables resulting in a post-audit dataset. We estimated the times from antiretroviral therapy initiation until death and first AIDS-defining event using the pre-audit data, the audit data, and the post-audit data.

RESULTS

The overall discrepancy rate between pre-audit and audit data (n = 250) across all audited variables was 17.1%. The estimated probability of mortality and an AIDS-defining event over time was higher in the audited data relative to the pre-audit data. Among patients represented in both the post-audit and pre-audit cohorts (n = 18,999), AIDS and mortality estimates also were higher in the post-audit data.

CONCLUSION

Though some changes may have occurred independently, our findings suggest that improved data quality following the audit may impact epidemiological inferences.

Aberrant natural killer (NK) cell activation and dysfunction among ART-treated HIV-infected adults in an African cohort

BACKGROUND

We examined NK cell phenotypes and functions after seven years of ART and undetectable viral loads (<50 copies/ml) with restored CD4 T-cell counts (≥500 cells/μl) and age-matched healthy-HIV-uninfected individuals from the same community.

METHODS

Using flow-cytometry, NK cell phenotypes were described using lineage markers (CD56^+/-CD16^+/-). NK cell activation was determined by expression of activation receptors (NKG2D, NKp44 and NKp46) and activation marker CD69. NK cell function was determined by CD107a, granzyme-b, and IFN-gamma production.

RESULTS

CD56 dim and CD56 bright NK cells were lower among ART-treated-HIV-infected than among age-matched-HIV-negative individuals; p = 0.0016 and p = 0.05 respectively. Production of CD107a (P = 0.004) and Granzyme-B (P = 0.005) was lower among ART-treated-HIV-infected relative to the healthy-HIV-uninfected individuals. NKG2D and NKp46 were lower, while CD69 expression was higher among ART-treated-HIV-infected than healthy-HIV-uninfected individuals.

CONCLUSION

NK cell activation and dysfunction persisted despite seven years of suppressive ART with "normalization" of peripheral CD4 counts.

Participation in a clinical trial of a text messaging intervention is associated with increased infant HIV testing: A parallel-cohort randomized controlled trial

Objective

Text messages significantly improve uptake of infant HIV testing in clinical trial contexts. Women who were excluded from a randomized trial in Kenya were followed to create a comparison between women who were enrolled and did not receive the study SMS intervention and women who were screened but not enrolled.

Design

Parallel-cohort randomized controlled trial analysis.

Methods

We compared time to infant HIV testing between women in three groups: the Trial SMS group, the Trial Control group, and the Comparison Cohort comprised of women who were screened but not enrolled.

Results

Of the 1,115 women screened, 388 (35%) were eligible for trial enrollment, and were randomized to receive either intervention text messages (Trial SMS; N = 195) or continue usual care (Trial Control; N = 193). Among 727 women not enrolled in the study (Comparison Cohort), we obtained infant HIV testing data from clinic records for 510 (70%). The cumulative probability of infant HIV testing was highest in the Trial SMS group (92.0%; 95% CI 87.5–95.3), followed by the Trial Control group (85.1%; 95% CI 79.5–89.8), and lowest among women in the Comparison Cohort (43.4%; 95% CI 39.2–47.8).

Conclusions

Both the Trial SMS group and the Trial Control group were significantly more likely to have their infants tested for HIV compared to the Comparison Cohort, providing evidence of a “clinical trial effect.” This analysis suggests that SMS interventions should be implemented as an adjunct to consistent and engaged delivery of basic health services.

Pediatric HIV care and treatment services in Tanzania: implications for survival

BACKGROUND

Improving child survival for HIV-infected children remains an important health agenda. We present progress regarding care and treatment services to HIV infected children in Tanzania.

METHODS

The National AIDS Control Programme Care and Treatment (CTC 2) database was used to obtain information of all children aged 0-14yearsenrolled in the HIV Care and Treatment Program between January 2011 and December 2014. We assessed eligibility for ART, time from enrolment to ART initiation, nutritional status, and mortality using Kaplan-Meier methods.

RESULTS

A total of 29,531 (14,304 boys and 15,227 girls) ART-naive children aged 0-14 years were enrolled during the period, approximately 6700 to 8000 children per year. The male to female ratio was 48:50. At enrolment 72% were eligible for ART, 2-3% of children were positive for TB, and 2-4% were severely malnourished. Between 2011 and 2014, 2368 (8%) died, 9243 (31%) were Lost to Follow-up and 17,920 (61%) were on care or ART. The probability of death was 31% (95% CI 26-35), 43% (40-47), 52% (49-55) and 61% (58-64) by 1,2, 5 and 10 years of age, respectively. The hazard of death was greatest at very young ages (<2 years old), and decreased sharply by 4 years old. Children who were on ART had around 10-15% higher survival over time.

CONCLUSIONS

Significant progress has been made regarding provision of paediatric HIV care and treatment in Tanzania. On average 7000 children are enrolled annually, and that approximately two thirds of children diagnosed under the age of 2 years were initiated on ART within a month. Provision of ART as soon as the child is diagnosed is the biggest factor in improving survival. However we noted that i) most children had advanced disease at the time of enrolment ii) approximately two-thirds of children were missing a baseline CD4 measurement and only 35% of children had either a CD4 count or percentage recorded, indicating limited access to CD4 testing services, and iii) 31% were lost to follow-up (LTFU). These challenges need to be addressed to improve early detection, enrolment and retention of HIV-infected children into care and improve documentation of services offered.

Maternal and newborn outcomes at a tertiary care hospital in Lusaka, Zambia, 2008-2012

OBJECTIVE

To measure key obstetric and neonatal outcomes recorded at a tertiary hospital in Zambia over a 5-year period.

METHODS

A retrospective analysis was conducted among women who had delivered at the University Teaching Hospital in Lusaka, between January 1, 2008, and December 31, 2012. Data were extracted from electronic medical records. The main outcomes were maternal mortality, cesarean delivery, prenatal or intrapartum hemorrhage, stillbirth, a 5-minute Apgar score of less than 7, and admission to the neonatal intensive care unit.

RESULTS

A total of 62 470 deliveries were recorded. Rates of maternal mortality, cesarean delivery, and hemorrhage during pregnancy all declined over time. Decreased admissions to the neonatal intensive care unit were observed; however, the rate spiked temporarily in late 2011 and early 2012 before returning to previous levels. The proportion of stillbirths remained stable over time but reports of a 5-minute Apgar score of less than 7 rose.

CONCLUSION

Routinely collected obstetric and neonatal data could aid ongoing program monitoring and should be used to guide quality improvement activities.

Outcomes in a Cohort of Patients Started on Antiretroviral Treatment and Followed up for a Decade in an Urban Clinic in Uganda

BACKGROUND

Short-medium term studies from sub-Saharan Africa show that, despite high early mortality, substantial loss to program, and high rates toxicity, patients on antiretroviral treatment have achieved outcomes comparable to those in developed settings. However, these studies were unable to account for long term outcomes of patients as they stayed longer on treatment.

OBJECTIVES

We aim to describe ten years outcomes of one of the first cohort of HIV positive patients started on antiretroviral treatment (ART) in Sub-Saharan Africa.

METHODS

We report 10-years outcomes including mortality, retention, CD4-count response, virological outcomes, ART regimens change from a prospective cohort of 559 patients initiating ART and followed up for 10 years Uganda.

RESULTS

Of 559 patients, 69.1% were female, median age (IQR) was 38 (33-44) years, median CD4-count (IQR) 98 (21-163) cell/μL; 74% were started on stavudine, lamivudine and nevirapine, 26% on zidovudine, lamivudine and efavirenz. After 10 years 361 (65%) patients were still in the study; 127 (22.7%) had died; 30 (5%) were lost to follow-up; 27 (5%) transferred; 18 (3%) withdrew consent. The probability of death was high in the first year (0.15, 95%, CI 0.12-0.18). The median CD4 count increased from 98 to 589 cell/μL (IQR: 450-739 cell/μL) with a median increase of 357 cells/μL (IQR: 128-600 cells/μL); 7.4% never attained initial viral suppression and of those who did 31.7% experienced viral failure. Three hundred and two patients had at least one drug substitution while on first line after a median of 40 months; 66 (11.9%) of the patients were switched to a second line PI-based regimen due to confirmed treatment failure.

CONCLUSIONS

Despite the high rate of early mortality due to advanced disease at presentation the outcomes from this cohort are encouraging, particularly the remarkable and incremental immune-recovery and a satisfactory rate of virologic suppression.

A structured approach to recording AIDS-defining illnesses in Kenya: A SNOMED CT based solution

INTRODUCTION

Several studies conducted in sub-Saharan Africa (SSA) have shown that routine clinical data in HIV clinics often have errors. Lack of structured and coded documentation of diagnosis of AIDS defining illnesses (ADIs) can compromise data quality and decisions made on clinical care.

METHODS

We used a structured framework to derive a reference set of concepts and terms used to describe ADIs. The four sources used were: (i) CDC/Accenture list of opportunistic infections, (ii) SNOMED Clinical Terms (SNOMED CT), (iii) Focus Group Discussion (FGD) among clinicians and nurses attending to patients at a referral provincial hospital in western Kenya, and (iv) chart abstraction from the Maternal Child Health (MCH) and HIV clinics at the same hospital. Using the January 2014 release of SNOMED CT, concepts were retrieved that matched terms abstracted from approach iii & iv, and the content coverage assessed. Post-coordination matching was applied when needed.

RESULTS

The final reference set had 1054 unique ADI concepts which were described by 1860 unique terms. Content coverage of SNOMED CT was high (99.9% with pre-coordinated concepts; 100% with post-coordination). The resulting reference set for ADIs was implemented as the interface terminology on OpenMRS data entry forms.

CONCLUSION

Different sources demonstrate complementarity in the collection of concepts and terms for an interface terminology. SNOMED CT provides a high coverage in the domain of ADIs. Further work is needed to evaluate the effect of the interface terminology on data quality and quality of care.

Trends in prevalence of diarrhoea, Kaposi's sarcoma, bacterial pneumonia, malaria and geohelminths among HIV positive individuals in Uganda

BACKGROUND

Trends in prevalence of opportunistic infections (OIs) associated with the human immunodeficiency virus (HIV) in resource poor settings have previously not been well documented. The objective of this study was to describe the trends in prevalence of Diarrhoea, Bacterial pneumonia, Kaposi's sarcoma, Malaria and Geohelminths among HIV positive individuals over a 12 year period in Uganda.

METHODS

Observation data for 5972 HIV positive individuals enrolled with the AIDS support organisation (TASO) in Uganda were analysed. Study participants were drawn from three HIV clinics located in different geographical areas of Uganda and followed from January 2002 to December 2013. The prevalence trends for the above OIs were plotted using the Box Jenkins moving average technique. X (2)-test for trend was used to test for the significance of the trends and Pearson's correlation coefficient used to test for the strength of linear relationship between OI prevalence and calendar time. Mixed effect linear regression was used to estimate average monthly change in prevalence with monthly variation modelled as a random effect.

RESULTS

A total of 204,871 monthly medical reports were retrieved and analysed. 73 % (4301/5972) were female with a median age of 32 years (inter-quartile range 26-39). Overall, significant decreasing mean annual prevalence trends (p < 0.05, X(2) trend) were observed for Diarrhoea (<1 month) with Pearson's correlation coefficient (r = -0.89), Malaria (r = -0.75), Bacterial Pneumonia (r = -0.52), and Geohelminth (r = -0.32). Non-significant increasing mean annual prevalence trend was observed for Kaposis sarcoma (p = 0.20, X(2) trend; r = +0.26). After adjusting for age, sex and clinic in a mixed effects linear regression model, average monthly prevalence declined significantly at a rate of 0.4 % for Kaposis sarcoma, 0.3 % for Geohelminths, 2 % for Malaria, 1 % for Bacterial Pneumonia and 3 % for Diarrhoea(<1 month). However, the rate of decline per month differed significantly (p < 0.05) by HIV clinic for Diarrhoea (<1 month), and age, sex and clinic for malaria.

CONCLUSIONS AND RECOMMENDATIONS

Overall, decreasing trends were observed in the above OIs. However the trends differed significantly by OI, geographical location and demographic characteristics. There is urgent need to integrate interventions targeting malaria and geohelminths in HIV programmes.

Trends in prevalence of selected opportunistic infections associated with HIV/AIDS in Uganda

Background

After more than a decade of establishing and expanding access to highly active antiretroviral therapy (HAART), empirical evidence on its impact on trends of opportunistic infections (OIs) associated with the deadly human immunodeficiency virus (HIV) in resource poor settings is scarce. The primary objective of this study was to assess the effect of HAART coverage on trends of five most common OIs in Uganda.

Methods

Observational data from January 2002 to December 2013 for 5972 HIV positive individuals attending the AIDS Support Organisation (TASO) HIV/AIDS care programme in Uganda were extracted and analysed. Trends were analysed using autoregressive moving average time series and mixed effects linear regression models adjusting for all available potential confounders.

Results

A total of 204,871 monthly medical reports were retrieved and analysed. Majority of the participants were female (73%) with a median age of 32 years (inter-quartile range 26–39). Overall, significant decreasing mean annual prevalence trends were observed for Mycobacterium tuberculosis, herpes zoster, genital ulcer and oral candidiasis (p < 0.05, X²_trend). Non-significant declining mean annual prevalence trend was observed for cryptococcal meningitis (p = 0.181, X²_trend). The largest impact of HAART was observed in Oral candidiasis and TB whose average annual prevalence reduced by 61% and 43% respectively following the introduction of HAART. Monthly series for TB, Herpes zoster and genital ulcers differed significantly by age and clinic but only genital ulcer series differed significantly by sex (p < 0.05, kruskal wallis). After controlling for the effects of age, sex and clinic (fixed) and monthly clustering (random effect) in a mixed effects linear regression model, all the five OIs showed a significant monthly change in prevalence (p < 0.001).

Conclusion

Overall, prevalence of most OIs declined especially after the introduction of HAART. However significant variations exist in the trends of different OIs in different geographical areas in Uganda. It is therefore important that site specific factors are properly identified to enable the development of targeted interventions.

Antiretroviral treatment response of HIV-infected children after prevention of mother-to-child transmission in West Africa

Introduction

We assessed the rate of treatment failure of HIV-infected children after 12 months on antiretroviral treatment (ART) in the Paediatric IeDEA West African Collaboration according to their perinatal exposure to antiretroviral drugs for preventing mother-to-child transmission (PMTCT).

Methods

A retrospective cohort study in children younger than five years at ART initiation between 2004 and 2009 was nested within the pWADA cohort, in Bamako-Mali and Abidjan-Côte d’Ivoire. Data on PMTCT exposure were collected through a direct review of children’s medical records. The 12-month Kaplan-Meier survival without treatment failure (clinical or immunological) was estimated and their baseline factors studied using a Cox model analysis. Clinical failure was defined as the appearance or reappearance of WHO clinical stage 3 or 4 events or any death occurring within the first 12 months of ART. Immunological failure was defined according to the 2006 World Health Organization age-related immunological thresholds for severe immunodeficiency.

Results

Among the 1035 eligible children, PMTCT exposure was only documented for 353 children (34.1%) and remained unknown for 682 (65.9%). Among children with a documented PMTCT exposure, 73 (20.7%) were PMTCT exposed, of whom 61.0% were initiated on a protease inhibitor-based regimen, and 280 (79.3%) were PMTCT unexposed. At 12 months on ART, the survival without treatment failure was 40.6% in the PMTCT-exposed group, 25.2% in the unexposed group and 18.5% in the children with unknown exposure status (p=0.002). In univariate analysis, treatment failure was significantly higher in children unexposed (HR 1.4; 95% CI: 1.0–1.9) and with unknown PMTCT exposure (HR 1.5; 95% CI: 1.2–2.1) rather than children PMTCT-exposed (p=0.01). In the adjusted analysis, treatment failure was not significantly associated with PMTCT exposure (p=0.15) but was associated with immunodeficiency (aHR 1.6; 95% CI: 1.4–1.9; p=0.001), AIDS clinical events (aHR 1.4; 95% CI: 1.0–1.9; p=0.02) at ART initiation and receiving care in Mali compared to Côte d’Ivoire (aHR 1.2; 95% CI: 1.0–1.4; p=0.04).

Conclusions

Despite a low data quality, PMTCT-exposed West African children did not have a poorer 12-month response to ART than others. Immunodeficiency and AIDS events at ART initiation remain the main predictors associated with treatment failure in this operational context.

Ontological specification of quality of chronic disease data in EHRs to support decision analytics: a realist review

This systematic review examined the current state of conceptualization and specification of data quality and the role of ontology based approaches to develop data quality based on "fitness for purpose" within the health context. A literature review was conducted of all English language studies, from January 2000-March 2013, which addressed data/information quality, fitness for purpose of data, used and implemented ontology-based approaches. Included papers were critically appraised with a "context-mechanism-impacts/outcomes" overlay. We screened 315 papers, excluded 36 duplicates, 182 on abstract review and 46 on full-text review; leaving 52 papers for critical appraisal. Six papers conceptualized data quality within the "fitness for purpose" definition. While most agree with a multidimensional definition of DQ, there is little consensus on a conceptual framework. We found no reports of systematic and comprehensive ontological approaches to DQ based on fitness for purpose or use. However, 16 papers used ontology-specified implementations in DQ improvement, with most of them focusing on some dimensions of DQ such as completeness, accuracy, correctness, consistency and timeliness. The majority of papers described the processes of the development of DQ in various information systems. There were few evaluative studies, including any comparing ontological with non-ontological approaches, on the assessment of clinical data quality and the performance of the application.

Reviewing progress: 7 year trends in characteristics of adults and children enrolled at HIV care and treatment clinics in the United Republic of Tanzania

BACKGROUND

To evaluate the on-going scale-up of HIV programs, we assessed trends in patient characteristics at enrolment and ART initiation over 7 years of implementation.

METHODS

Data were from Optimal Models, a prospective open cohort study of HIV-infected (HIV+) adults (≥15 years) and children (<15 years) enrolled from January 2005 to December 2011 at 44 HIV clinics in 3 regions of mainland Tanzania (Kagera, Kigoma, Pwani) and Zanzibar. Comparative statistics for trends in characteristics of patients enrolled in 2005-2007, 2008-2009 and 2010-2011 were examined.

RESULTS

Overall 62,801 HIV + patients were enrolled: 58,102(92.5%) adults, (66.5% female); 4,699(7.5%) children.Among adults, pregnant women enrolment increased: 6.8%, 2005-2007; 12.1%, 2008-2009; 17.2%, 2010-2011; as did entry into care from prevention of mother-to-child HIV transmission (PMTCT) programs: 6.6%, 2005-2007; 9.5%, 2008-2009; 12.6%, 2010-2011. WHO stage IV at enrolment declined: 27.1%, 2005-2007; 20.2%, 2008-2009; 11.1% 2010-2011. Of the 42.5% and 29.5% with CD4+ data at enrolment and ART initiation respectively, median CD4+ count increased: 210 cells/μL, 2005-2007; 262 cells/μL, 2008-2009; 266 cells/μL 2010-2011; but median CD4+ at ART initiation did not change (148 cells/μL overall). Stavudine initiation declined: 84.9%, 2005-2007; 43.1%, 2008-2009; 19.7%, 2010-2011.Among children, median age (years) at enrolment decreased from 6.1(IQR:2.7-10.0) in 2005-2007 to 4.8(IQR:1.9-8.6) in 2008-2009, and 4.1(IQR:1.5-8.1) in 2010-2011 and children <24 months increased from 18.5% to 26.1% and 31.5% respectively. Entry from PMTCT was 7.0%, 2005-2007; 10.7%, 2008-2009; 15.0%, 2010-2011. WHO stage IV at enrolment declined from 22.9%, 2005-2007, to 18.3%, 2008-2009 to 13.9%, 2010-2011. Proportion initiating stavudine was 39.8% 2005-2007; 39.5%, 2008-2009; 26.1%, 2010-2011. Median age at ART initiation also declined significantly.

CONCLUSIONS

Over time, the proportion of pregnant women and of adults and children enrolled from PMTCT programs increased. There was a decline in adults and children with advanced HIV disease at enrolment and initiation of stavudine. Pediatric age at enrolment and ART initiation declined. Results suggest HIV program maturation from an emergency response.

Prevalence and burden of diseases presenting to a general pediatrics ward in Gondar, Ethiopia

BACKGROUND

Little is known about pediatric hospital admissions in Ethiopia.

METHODS

This cross-sectional study analyzed all data entered into the Gondar University Hospital pediatric ward's admission registration books over 1 year. Patient age, sex, origin, length of stay, diagnosis and discharge condition were transcribed into an electronic database for all observations. Missing data were retrieved by chart and death certificate review. Primary outcome measures included death and death in the first 24 h of admission.

RESULTS

In all, 1927 patients were admitted to our facility during the year of study. Of these, 64.5% improved, 4.6% were discharged unchanged, 6.5% disappeared and 7.5% died; the remaining 17.0% of outcome data were registered as 'non-death' but could not be specified further. The median age of admission was 2.2 years (interquartile range 1-7 years), with more admissions for children younger than 5 years (70.3%) and more male subjects admitted than female subjects (59.6% male). The median length of stay was 4.0 days (interquartile range 2-10 days). Eighty-one percent of admissions originated from Gondar or its neighboring districts. Most admissions carried a respiratory, nutritional or infectious diagnosis (47.5, 46.8 and 36.5%, respectively). Conditions diagnosed most commonly (>200 cases) included community-acquired pneumonia (812 cases), severe acute malnutrition (381), anemia (274) and acute gastroenteritis (219). Seven diagnoses were associated with mortality after adjusting for demographic covariates: severe acute malnutrition (odds ratio (OR) 2.5, P < 0.001), coma (OR 4.2, P < 0.001), meningitis (OR 2.3, P = 0.018), congestive heart failure (OR 2.4, P = 0.001), severe dehydration (OR 2.5, P = 0.004), aspiration pneumonia (OR 5.4, P < 0.001) and sepsis (OR 3.2, P < 0.001). Thirty-three percent of deaths occurred in the first 24 h of admission, with four diagnoses associated with first-24-h mortality after adjusting for demographic covariates: coma (OR 7.0, P < 0.001), meningitis (OR 3.2, P = 0.008), congestive heart failure (OR 3.1, P = 0.008) and aspiration pneumonia (OR 12.1, P < 0.001).

CONCLUSIONS

This study demonstrates a mortality pattern at our hospital that differs considerably from Ethiopia as a whole, and may differ from other hospitals in sub-Saharan Africa. Hospitals must look beyond national and regional agenda when identifying mortality reduction targets.

Scaling up pediatric HIV care and treatment in Africa: clinical site characteristics associated with favorable service utilization

BACKGROUND

To improve pediatric enrollment and retention in HIV treatment programs in Africa, we examined factors associated with service utilization within the Elizabeth Glaser Pediatric AIDS Foundation program in Côte d'Ivoire, Mozambique, South Africa, Tanzania and Zambia.

METHODS

We retrospectively reviewed characteristics of clinical sites providing HIV treatment services within our program. For each site, favorable pediatric program outcomes were defined as a cumulative number or percentage of pediatric enrollment in care or antiretroviral therapy (ART) more than the pooled median value or an attrition rate less than 10%. We compared proportions of sites with favorable outcomes among those with or without selected characteristics. Adjusted odds ratios (aORs) and 95% confidence intervals (CIs) were determined using logistic regression analyses, accounting for potential confounding factors.

RESULTS

Over 4 years, 33,331 children were enrolled, including 18,255 on ART, across 220 sites. Characteristics associated with favorable pediatric enrollment were nutritional support (aOR = 8.9; CI: 2.8 to 28.4), linkages with associations of people living with HIV (aOR = 4.2; CI: 1.8 to 9.5), early infant diagnosis (aOR = 3.3; CI: 1.5 to 7.1), and on-site prevention of mother-to-child transmission services (aOR = 3.1; CI: 1.0 to 11.1). Similarly, linkages with people living with HIV, early infant diagnosis, and prevention of mother-to-child transmission were associated with high proportion of children on ART younger than 2 years of age. Home-based care was associated with low pediatric attrition rates (aOR = 2.9; CI: 1.4 to 5.8).

CONCLUSIONS

Certain site characteristics were associated with favorable pediatric enrollment and retention in our program. Expanding these characteristics to improve pediatric HIV treatment in Africa warrants further evaluation.

Implementation of provider-based electronic medical records and improvement of the quality of data in a large HIV program in Sub-Saharan Africa

Introduction

Starting in June 2010 the Infectious Diseases Institute (IDI) clinic (a large urban HIV out-patient facility) switched to provider-based Electronic Medical Records (EMR) from paper EMR entered in the database by data-entry clerks. Standardized clinics forms were eliminated but providers still fill free text clinical notes in physical patients’ files. The objective of this study was to compare the rate of errors in the database before and after the introduction of the provider-based EMR.

Methods and Findings

Data in the database pre and post provider-based EMR was compared with the information in the patients’ files and classified as correct, incorrect, and missing. We calculated the proportion of incorrect, missing and total error for key variables (toxicities, opportunistic infections, reasons for treatment change and interruption). Proportions of total errors were compared using chi-square test. A survey of the users of the EMR was also conducted. We compared data from 2,382 visits (from 100 individuals) of a retrospective validation conducted in 2007 with 34,957 visits (from 10,920 individuals) of a prospective validation conducted in April–August 2011. The total proportion of errors decreased from 66.5% in 2007 to 2.1% in 2011 for opportunistic infections, from 51.9% to 3.5% for ART toxicity, from 82.8% to 12.5% for reasons for ART interruption and from 94.1% to 0.9% for reasons for ART switch (all P<0.0001). The survey showed that 83% of the providers agreed that provider-based EMR led to improvement of clinical care, 80% reported improved access to patients’ records, and 80% appreciated the automation of providers’ tasks.

Conclusions

The introduction of provider-based EMR improved the quality of data collected with a significant reduction in missing and incorrect information. The majority of providers and clients expressed satisfaction with the new system. We recommend the use of provider-based EMR in large HIV programs in Sub-Saharan Africa.

Towards an ontology for data quality in integrated chronic disease management: a realist review of the literature

PURPOSE

Effective use of routine data to support integrated chronic disease management (CDM) and population health is dependent on underlying data quality (DQ) and, for cross system use of data, semantic interoperability. An ontological approach to DQ is a potential solution but research in this area is limited and fragmented.

OBJECTIVE

Identify mechanisms, including ontologies, to manage DQ in integrated CDM and whether improved DQ will better measure health outcomes.

METHODS

A realist review of English language studies (January 2001-March 2011) which addressed data quality, used ontology-based approaches and is relevant to CDM.

RESULTS

We screened 245 papers, excluded 26 duplicates, 135 on abstract review and 31 on full-text review; leaving 61 papers for critical appraisal. Of the 33 papers that examined ontologies in chronic disease management, 13 defined data quality and 15 used ontologies for DQ. Most saw DQ as a multidimensional construct, the most used dimensions being completeness, accuracy, correctness, consistency and timeliness. The majority of studies reported tool design and development (80%), implementation (23%), and descriptive evaluations (15%). Ontological approaches were used to address semantic interoperability, decision support, flexibility of information management and integration/linkage, and complexity of information models.

CONCLUSION

DQ lacks a consensus conceptual framework and definition. DQ and ontological research is relatively immature with little rigorous evaluation studies published. Ontology-based applications could support automated processes to address DQ and semantic interoperability in repositories of routinely collected data to deliver integrated CDM. We advocate moving to ontology-based design of information systems to enable more reliable use of routine data to measure health mechanisms and impacts.

The effect of electronic medical record-based clinical decision support on HIV care in resource-constrained settings: a systematic review

BACKGROUND

It is estimated that one million people infected with HIV initiate anti-retroviral therapy (ART) in resource-constrained countries annually. This occurs against a background of overburdened health workers with limited skills to handle rapidly changing treatment standards and guidelines hence compromising quality of care. Electronic medical record (EMR)-based clinical decision support systems (CDSS) are considered a solution to improve quality of care. Little evidence, however, exists on the effectiveness of EMR-based CDSS on quality of HIV care and treatment in resource-constrained settings.

OBJECTIVE

The aim of this systematic review was to identify original studies on EMR-based CDSS describing process and outcome measures as well as reported barriers to their implementation in resource-constrained settings. We characterized the studies by guideline adherence, data and process, and barriers to CDSS implementation.

METHODS

Two reviewers independently assessed original articles from a search of the MEDLINE, EMBASE, CINAHL and Global Health Library databases until January 2012. The included articles were those that evaluated or described the implementation of EMR-based CDSS that were used in HIV care in low-income countries.

RESULTS

A total of 12 studies met the inclusion criteria, 10 of which were conducted in sub-Saharan Africa and 2 in the Caribbean. None of the papers described a strong (randomized controlled) evaluation design. Guideline adherence: One study showed that ordering rates for CD4 tests were significantly higher when reminders were used. Data and process: Studies reported reduction in data errors, reduction in missed appointments, reduction in missed CD4 results and reduction in patient waiting time. Two studies showed a significant increase in time spent by clinicians on direct patient care. Barriers to CDSS implementation: Technical infrastructure problems such as unreliable electric power and erratic Internet connectivity, clinicians' limited computer skills and failure by providers to comply with the reminders are key impediments to the implementation and effective use of CDSS.

CONCLUSION

The limited number of evaluation studies, the basic and heterogeneous study designs, and varied outcome measures make it difficult to meaningfully conclude on the effectiveness of CDSS on quality of HIV care and treatment in resource-limited settings. High quality evaluation studies are needed. Factors specific to implementation of EMR-based CDSS in resource-limited setting should be addressed before such countries can demonstrate its full benefits. More work needs to be done to overcome the barriers to EMR and CDSS implementation in developing countries such as technical infrastructure and care providers' computer illiteracy. However, simultaneously evaluating and describing CDSS implementation strategies that work can further guide wise investments in their wider rollout.

Improving data quality and supervision of antiretroviral therapy sites in Malawi: an application of Lot Quality Assurance Sampling

Background

High quality program data is critical for managing, monitoring, and evaluating national HIV treatment programs. By 2009, the Malawi Ministry of Health had initiated more than 270,000 patients on HIV treatment at 377 sites. Quarterly supervision of these antiretroviral therapy (ART) sites ensures high quality care, but the time currently dedicated to exhaustive record review and data cleaning detracts from other critical components. The exhaustive record review is unlikely to be sustainable long term because of the resources required and increasing number of patients on ART. This study quantifies the current levels of data quality and evaluates Lot Quality Assurance Sampling (LQAS) as a tool to prioritize sites with low data quality, thus lowering costs while maintaining sufficient quality for program monitoring and patient care.

Methods

In January 2010, a study team joined supervision teams at 19 sites purposely selected to reflect the variety of ART sites. During the exhaustive data review, the time allocated to data cleaning and data discrepancies were documented. The team then randomly sampled 76 records from each site, recording secondary outcomes and the time required for sampling.

Results

At the 19 sites, only 1.2% of records had discrepancies in patient outcomes and 0.4% in treatment regimen. However, data cleaning took 28.5 hours in total, suggesting that data cleaning for all 377 ART sites would require over 350 supervision-hours quarterly. The LQAS tool accurately identified the sites with the low data quality, reduced the time for data cleaning by 70%, and allowed for reporting on secondary outcomes.

Conclusions

Most sites maintained high quality records. In spite of this, data cleaning required significant amounts of time with little effect on program estimates of patient outcomes. LQAS conserves resources while maintaining sufficient data quality for program assessment and management to allow for quality patient care.

Strengthening health systems through HIV monitoring and evaluation in Sub-Saharan Africa

PURPOSE OF REVIEW

In this review, we describe the challenges faced by using clinical cohorts to perform Monitoring and Evaluation (M&E) and the possible solutions for increasing and strengthening health systems in low-income and middle-income countries.

RECENT FINDINGS

HIV scale-up has facilitated the transition from paper-based medical records to electronic medical records at hundreds of sites in most of the countries in sub-Saharan Africa and the implementation of national HIV databases. However, National M&E systems in resource-limited settings tend to be chronically challenged by persistently incomplete reporting and inaccurate data, which undermines their usefulness. In low-income and middle-income countries, new information technologies such as Web-based systems and mobile phone networks are expanding rapidly. Their use will improve data quality and, therefore, reduce participant dropout and improve reporting rates. These systems have the potential to allow real time access to summary reports and to integrate data from other settings such as maternal and infant health clinics.

SUMMARY

The efforts to address the AIDS pandemic in sub-Saharan Africa will require enhanced information systems. Stronger systems are needed to deliver primary healthcare for those with and without HIV, and it is also essential to build and take advantage of synergies across health information systems.