EVALUATION OF MEDICATION SAFETY DATA-1.pptx

Crpv

1. CRPV STATISTICAL METHODS FOR EVALUATING SAFETY MEDICATION DATA PRESENTED BY, A.CHIFORA, M.Pharm-2 nd SEMESTER, DEPARTMENT OF PHARMACOLOGY, COP-SRIPMS

2. INTRODUCTION DESCRIPTIVE HYPOTHESIS TIME-TO-EVENT REGRESSION DISPROPORTIONALITY CONCLUSION REFERENCES

3. INTRODUCTION Statistical methods for evaluating medication safety data differ depending on the stage of the drug's life cycle. DURING CLINICAL TRIALS POST-MARKETING SURVEILLANCE 🐭Descriptive Statistics. 🐭Hypothesis Testing 🐭Time-to-event analysis 🐭Regression Analysis 👩🏻‍🔬Disproportionality Analysis 👩🏻‍🔬Bayesian Methods

4. DURING CLINICAL TRIALS

5. DESCRIPTIVE STATISTICS 🌀Side effect Reports = 100 patients Headache 20% (20 patients) Diarrhoea  15% (15 patients) Dizziness10% (10 patients) Nausea5% (5 patients) Most common: Headache 🌀Age of patients with dizziness Mean value= 54.6 years Median value= 55 years Mean value  Average Median value Middle value 🌀Range of years: 62-45=17 years Range:17 years Age spread: 17 years Descriptive statistics provide a simple, initial understanding of a drug’s safety profile

6. HYPOTHESIS TESTING STEP-1: FORMULATE HYPOTHESIS NULL HYPOTHESIS(H ): Drug X Adverse event ALTERNATIVE HYPOTHESIS(H a ): Drug X Adverse event STEP-2: COLLECT DATA & CALCULATE Clinical Trial Data Statistical Test (e.g. Chi-square test) GROUP PATIENTS ADVERSE EVENT Drug X 1000 50 (5%) Placebo 1000 20(2%) P-value=0.05 P-value  <0.05  Adverse event is produced by the drug P-value >0.05  Adverse event is not produced by the drug

7. CHI-SQUARED TEST

8. CONTD..

9. CONTD.. Ho  Drug A does not cause headache

10. TIME-TO-EVENT ANALYSIS Kaplan-Meier Curve ⏰Time-to-event analysis is also known as survival analysis ⏰Represented by Kaplan-Meier curve ⏰GOAL: To estimate the probability of survival (free from adverse effects) ⏰Y-AXIS: Cumulative probability of being event free  defines the dependent variable and its scale ⏰Median Survival time is employed; Time where the curve crosses 50% survival probability line

11. Drug dose (mg) Adverse Event Occurrence Regression line quantifies the linear association between drug dose and adverse event occurrence REGRESSION ANALYSIS 🌸Regression analysis draws a line through the data points to show the overall trend. This line helps predict how likely it is for adverse events to happen at different drug doses. 🌸By using this method, we can spot which doses carry higher risks 🌸Regression analysis uses math to show and predict how adverse effects changes as the drug dose increases.

12. POST-MARKETING SURVEILLANCE

13. DISPROPORTIONATE ANALYSIS A+OD 100 NO Disproportionality signal ✅ ✅ Observed = 100 ✅ Expected = 100 ✅ Observed = Expected ✅ NO DIFFERENCE! Disproportionality signal Detected ❌ A+OD 250 ❌ Observed = 250 ❌ Expected = 100 ❌ Observed  Expected ❌ BIG DIFFERENCE!

14. BAYESIAN METHOD Allow us to update our knowledge about a drug as new data comes in! 📜 Starting point 📜Represents everything about the drug’s safety before analysing new reports 🔍 This is the new, objective evidence  often from a clinical trial or from spontaneous adverse event reporting system that challenges the prior belief This is the result. 📜+🔍= ✅ ✅The Bayesian analysis mathematically combines the Prior belief with the new data to get the most accurate and updated assessment of risk

15. SIGNIFICANCE OF STATISTICAL METHODS IN EVALUATING MEDICATION SAFETY DATA Safety signal Detection Risk Quantification and Comparison Controlling for confounding factors Establishing Statistical significance Characterizing event timing Regulatory compliance Analysing rare events I MPOR T ANT

16. RECENTLY EMPLOYED ADVANCED STATISTICAL METHODS 👍Bayesian Confidence Propagation Neural Network 👍Machine learning methods 👍Propensity score matching and casual inferences 👍Meta analysis 👍Safety Visualization

17. CONCLUSION 🤝Statistics are the indispensable bridge between data and decision 🤝They ensure early detection of unknown risks 🤝They validate risk assessments 🤝Never guessing, always quantifying! 🤝Focus on the patient 🤝Silent guardians of the public health

18. REFERENCES Kaladharan S. Statistical methods for Pharmacovigilance. Overview of statistical methods used in pharmacovigilance[Internet]. India; [ cited 2025 Sept 19]. Available from: https://share.google/Gq9xJCBNDnyCRI1fM Xia A, Jiang Q. Statistical evaluation of drug safety data [Internet]. Global biostatistical science; [cited 2025 Sept 19]. Available from: https://share.google/jCIsmVEk3nJXmvIlr Kim HR, Sung MD, Park JA, Jeong K, Kim HH, Lee S, et al .. Analysing adverse drug reaction using statistical and machine learning methods [Internet]. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC9276413/ Sam K. Statistical methods [Internet]. India; [cited 2025 Sept 19 ]. Available from: https://www.scribd.com/presentation/880142507 Denfeld QE, Burger D, Lee CS. Survival analysis 101: an easy start guide to analyzing time-to-event data.[Internet]. [cited 2025 Sept 28]. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC927641/