We conducted an assessment of knowledge about blood borne pathogens (BBP) and use of universal precautions at first level care facilities (FLCF) in two districts of Pakistan.

Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author). Background: Despite the availability of free routine immunizations in low and middle-income countries (LMICs), many children are incompletely vaccinated, vaccinated late for age or drop out over the course of the immunization schedule. Without the technology to model and visualize risk on large datasets, vaccinators and policy-makers are unable to identify and target groups and individuals at high risk of dropping out thus default rates remain high preventing universal immunization coverage. Predictive analytics algorithm leverages Artificial Intelligence (AI), and uses statistical modeling, machine learning and multi-dimensional data mining to accurately identify children who are most likely to delay or miss their follow-up immunization visits. Objective: To conduct feasibility testing and validation of a predictive analytics algorithm to identify which children are likely to miss the follow-up vaccines through risk profiling of clients. Methods: The algorithm was developed using 47,554 longitudinal immunization records, split into a training and validation cohort. Four machine learning models (random forest, recursive partitioning, support vector machines and C-forest) were used to generate the algorithm to predict, for each child, the likelihood of defaulting on the follow-up immunization visit. The models were encapsulated in the predictive engine, which decided the most appropriate method to apply in a given case. Each of the models was assessed on: accuracy, precision (positive predictive value), sensitivity, specificity and negative predictive value, and area under curve. Results: Out of 11,889 cases in validation dataset, the random forest model correctly predicted 8,994 cases yielding 94.9% sensitivity and 54.9% specificity. The C-Forest Model, Support Vector Machines and Recursive Partitioning Models improved prediction respectively by achieving 352, 376 and 389 visits above the predictions made by the random forest model. All models had a C-statistic of 0.750 or above while the highest statistic (AUC 0.791, 95% CI: 0.784– 0.798) was seen for the Recursive Partitioning algorithm. Conclusions: This feasibility study demonstrates predictive analytics can accurately identify children who are at higher risk for defaulting on follow-up immunization visits. Correct identification of potential defaulters opens a window for evidence-based targeted interventions in resource limited settings to achieve optimal immunization coverage and timeliness.

Objectives: To estimate the prevalence of zero dose children (who have not received any dose of pentavalent (diphtheria, tetanus, pertussis, Haemophilus influenzae type B and hepatitis B) vaccine by their first birthday) among those who interacted with the immunisation system in Sindh, Pakistan along with their sociodemographic characteristics and risk factors. Design and participants: We conducted a descriptive analysis of child-level longitudinal immunisation records of 1 467 975 0–23 months children from the Sindh’s Zindagi Mehfooz (Safe Life) Electronic Immunisation Registry (ZM-EIR), for the birth cohorts of 2017 and 2018. Setting: Sindh province, Pakistan which has a population of 47.9 million people and an annual birth cohort of 1.7 million. Primary and secondary outcome measures: The primary outcome measure was zero dose status among enrolled children. Logistic regression was performed to identify the risk factors associated with the zero dose status. Results: Out of 1 467 975 children enrolled in the ZM-EIR in Sindh, 10.6% (154 881/1 467 975) were zero dose. There were sharp inequities across the 27 districts. Zero dose children had a lower proportion of hospital births (28.5% vs 34.0%; difference 5.5 percentage points (pp) (95% CI 5.26 to 5.74); p<0.001) and higher prevalence from slums (49.5% vs 42.3%; difference 7.2 pp (95% CI 6.93 to 7.46); p<0.001), compared with non-zero dose children. Children residing in urban compared with rural areas were at a higher risk (relative risk (RR): 1.20; p<0.001; 95% CI 1.18 to 1.22), while children with educated compared with uneducated mothers were at a lower risk of being zero dose (RR: 0.47–0.96; p<0.001; 95% CI 0.45 to 0.98). Conclusions: Despite interacting with the immunisation system, 1 out of 10 children enrolled in the ZM-EIR in Sindh were zero dose. It is crucial to monitor the prevalence of zero dose children and investigate their characteristics and risk factors to effectively reach and follow-up with them.