Maternal Health Measurements Moussa LY MPH Monitoring Evaluation

Maternal Health Measurements Moussa LY , MPH Monitoring & Evaluation Specialist Maternal Health/Family Planning Project Dakar - Senegal

Outline course • To determine causes related to maternal health complications • To identify best indicators (Independent variables associated with maternal health complications) • To use epidemiology approaches to test the association between maternal health complications and suspected causes.

Delivery of Improved Services for Health Project • Aims related to reproductive and maternal health: – increase service availability – improve the quality of services – increase public knowledge, change attitudes and behavior – increase sustainability of services

Timeline for Evaluation Activities PERIOD OF THE EVALUATION Project Implementation Populationbased surveys Facilitybased surveys Phase II KABS ABS HFR, CEI, HPI HFR, CEI, OHPC HFR, OHPC

Measurements and Analysis: 1. Data collection • Questionnaire to administer to community members (sisters method), exit clients, health providers • Medical records from health facilities • Focus group discussion with communities • Observation of Health providers and clients inter-action

Measurements and Analysis: 2. Selected variables DEPENDANT VARIABLE • Number of Antenatal Care clients with complications or those who died INDEPENDANT VARIABLES • Socio-demographic factors (Age, Marital status, Level of education, place of residence) • Nutrition status • Reproductive Health history (Parity, Abortions, Age at the first sexual intercourse, Age at the first birth, Unwanted pregnancies, etc…) • Previous diseases • Complications related to the diseases

Measurements and Analysis: 3. Measures of Association 2 • Chi square ( ) test with = 0. 05 and P the level of significance – 2 tabulate = 2 – 2 tabulate > 2 • Hypotheses – H 0: No significant association between Maternal health complications with independent variables – H 1: Existence of significant association between Maternal health complications with independent variables • Decision rule – Reject H 0 if P < – Accept H 0 if P ≥

Measurements and Analysis: 3. Measures of Association (Cont. ) • Odd ratio: The ratio of the odds of a condition in the exposed compared with the odds of the condition in the unexposed. • Decision rule: – OR = 1: No association – OR > 1: Diseases are risk factors – OR < 1: Diseases are protective

Measurements and Analysis: 3. Measures of Association (Cont) • Expression of misclassification in terms of sensitivity and specificity of Knowing the likelihood of maternal health complications with a positive test and in those with negative test. • Sensitivity is the probability of a test given that a woman with one of the independent variables has health complications • Specificity is the probability of a negative test given • Receiver Operating Characteristic (ROC) curve is a plot of the true positive rate against false positive rate for the different possible cut off points of a diagnostic test

Interpretation of the ROC curve

Normal Probability Plot

Logistic Regression • Logistic regression model is simply a non linear transformation of the linear regression. Binary logistic regression is a type of regression analysis where the dependant variable (maternal death) is dummy variable (coded 0, 1) – 1: When the woman has health complications – 0: When the woman has not health complications

References • Management Sciences for Health, Evaluation Handbook, Dec. 2003 • Measure Evaluation, Guideline for Family Planning Monitoring and Evaluation, 2000 • UNAIDS, Guideline for AIDS monitoring and Evaluation, 2000 • USAID, Management of Reproductive, Maternal and Child Health Project, 2000