Pharm D Year 5, 5.2 Pharmacoepidemiology and Pharmacoeconomics — Pharm D Year 5

Unit 1: Pharmacoepidemiology, Scope and Outcomes

Syllabus scope

This unit covers the definition, origin, evolution, need, aims, applications, and measurement of outcomes in pharmacoepidemiology.

Pharmacoepidemiology applies epidemiologic reasoning to medicine use and effects in populations. It connects pharmacology, clinical care, statistics, public health, and health-system data. Randomized trials provide controlled evidence, but population studies are needed to examine rare harms, long latency, effectiveness, adherence, diverse patients, utilization, and routine practice.

Pharmacoepidemiology cycle

1. Define the medicine, population, exposure, comparison, and outcome.
2. Select a valid data source and design.
3. Measure use, events, risk, and uncertainty.
4. Address confounding, bias, missingness, and time.
5. Interpret clinical and public-health meaning.
6. Communicate action, limitations, and further monitoring.

Outcome measures

Outcomes may be clinical, humanistic, or economic. Clinical outcomes include disease, hospitalization, death, laboratory change, adverse reaction, or treatment success. Humanistic outcomes include symptoms, function, quality of life, satisfaction, and adherence. Economic outcomes include medicine cost, service use, productivity, and total care cost.

Drug-use measures include prescriptions, units dispensed, defined daily doses, prescribed daily doses, days of therapy, persistence, switching, and discontinuation. Definitions must specify the denominator, observation period, product, strength, route, and population.

Clinical correlation

A high number of prescriptions may reflect a large population, repeated refills, or inappropriate use. A low event count may reflect under-detection. The pharmacist should ask what the measure represents before drawing a conclusion.

Worked example

An antibiotic-use audit reports 1,000 prescriptions. Without population size, duration, indication, and repeat dispensing, it cannot establish excessive use. A rate per population or course-days measure may be more informative.

FAQs

Is pharmacoepidemiology only about adverse reactions? No. It studies use, benefit, harm, adherence, outcomes, and policy.

Why define the denominator? The same numerator can mean different things in different populations.

Takeaways

• Study medicine use in real populations with explicit definitions.
• Measure clinical, humanistic, and economic outcomes.
• Interpret counts only with population, time, and exposure context.

Integrated practice note

Pharmacoepidemiology is useful because medicines behave differently across real populations than in a selected trial sample. Patients vary in disease severity, adherence, co-morbidity, access, genetics, prescribing, and follow-up. A population signal should therefore be translated into a clinical question: who is affected, how large is the absolute problem, what alternatives exist, and what action can reduce harm without removing effective treatment unnecessarily? Good studies make those links explicit.

They also distinguish effectiveness from efficacy, identify who was not observed, and report whether the result is transferable to another setting. A population average can conceal differences by age, disease severity, access, or treatment history. Subgroup findings require cautious interpretation when sample size or prespecification is limited. The study question should always state whether it concerns use, benefit, harm, or value.

References

1. PharmacoLearn, Pharm D Year 5 syllabus: https://pharmacolearn.fyi/syllabus/pharm-d/year-5/
2. Pharmacy Council of India, Pharm.D. Regulations, 2008.
3. Strom, Kimmel and Hennessy, Textbook of Pharmacoepidemiology, latest available edition.

Unit 2: Incidence, Prevalence, Drug Use and Adherence

Syllabus scope

This unit covers prevalence, incidence, incidence rate, monetary units, prescriptions, dispensed units, defined daily doses, prescribed daily doses, and medication-adherence measurement.

Measures

Prevalence is the proportion of a population with a condition at a point or period. Incidence proportion describes new events among people at risk over a defined period. Incidence rate uses person-time in the denominator. The choice depends on whether the population is closed, follow-up differs, and the question concerns existing burden or new occurrence.

Defined daily dose is a standardized assumed average maintenance dose for a medicine’s main indication in adults. It is not necessarily the prescribed dose or an individual patient’s correct dose. Prescribed daily dose reflects actual prescribing and may differ from the defined daily dose.

Measurement sequence

1. Define numerator, denominator, population at risk, and time.
2. Specify medicine, strength, route, and indication.
3. Select prescription, dispensing, administration, or patient-reported data.
4. Check repeat records, switching, death, migration, and missingness.
5. Report rate with uncertainty and limitations.

Adherence

Adherence has initiation, implementation, and persistence dimensions. Measures include self-report, pill count, refill records, electronic monitoring, drug levels, and clinical response. Each measures a different aspect and can be biased. Medication possession ratio and proportion of days covered require clearly defined observation windows and handling of early refills or stockpiling.

Clinical correlation

A patient may possess medicine without taking it, take it differently, or stop because of adverse effects or cost. A low refill measure is a signal for discussion, not proof of intentional nonadherence.

Worked example

If a patient has medicine available for 80 of 100 observed days, a simple possession proportion is 0.80, but interpretation depends on dose, early supply, hospitalization, sample packs, and whether the medicine is intended for daily use.

FAQs

Is defined daily dose a patient dose? No.

Which adherence measure is best? It depends on the behavior and question; triangulation is often stronger.

Takeaways

• Use incidence, prevalence, and incidence rate for different questions.
• Distinguish DDD, PDD, prescriptions, and administration.
• Treat adherence measures as indicators with limitations.

Integrated practice note

Measurement choices can change the apparent result. A prescription count measures ordering, a dispensing count measures supply, an administration record measures recorded delivery, and a patient interview measures reported use. None alone proves ingestion or therapeutic effect. Define the behavior of interest, select the closest feasible measure, and triangulate with clinical response, refill history, interviews, and barriers when interpreting adherence.

Adherence is dynamic. Initiation may fail immediately, implementation may vary day to day, and persistence may end after adverse effects, cost, improvement, or changing beliefs. Interventions should target the identified barrier, such as simplification, education, access support, reminder systems, side-effect management, or shared decision-making. Follow-up should measure both behavior and the clinical outcome it was intended to improve. The same measure should be interpreted consistently across groups and time.

References

1. PharmacoLearn, Pharm D Year 5 syllabus: https://pharmacolearn.fyi/syllabus/pharm-d/year-5/
2. Pharmacy Council of India, Pharm.D. Regulations, 2008.
3. World Health Organization, adherence to long-term therapies resources.

Unit 3: Risk Measurement in Pharmacoepidemiology

Syllabus scope

This unit covers risk, attributable risk, relative risk, time-risk relationships, and odds ratios.

Risk measures

Absolute risk is the probability of an event over a defined period. Relative risk compares risks in exposed and unexposed groups. Attributable risk is the excess risk associated with exposure under the comparison assumptions. Odds ratio compares odds and can approximate relative risk when the outcome is uncommon, but it can exaggerate the apparent effect when the outcome is common.

Risk-analysis sequence

1. Define exposure and comparison groups.
2. Establish who is at risk and when follow-up begins.
3. Identify event definition and ascertainment method.
4. Calculate absolute and relative measures.
5. Assess confidence intervals, confounding, and bias.
6. Translate the result into clinical and population meaning.

Time-risk relationships matter because exposure, induction period, latency, depletion of susceptibles, treatment changes, and censoring alter risk. Immortal-time bias can occur when a period is incorrectly assigned to an exposed group during which the event could not have occurred under the study definition.

Worked example

If an event occurs in 10 of 1,000 exposed people and 5 of 1,000 unexposed people, risks are 1% and 0.5%, relative risk is 2, and attributable risk is 0.5 percentage points. The absolute increase is small but may matter if the event is severe.

Clinical correlation

A relative risk without baseline risk can mislead decisions. A pharmacist should ask about event severity, preventability, treatment alternatives, and patient preferences.

FAQs

Does relative risk prove causation? No. Confounding and bias can create or distort associations.

Why specify time? Risk is meaningless without an observation period.

Takeaways

• Report absolute and relative effects together.
• Define time at risk and event ascertainment.
• Interpret odds ratios carefully when outcomes are common.

Integrated practice note

Risk communication should include both relative and absolute information. A doubled risk may represent an increase from one in ten thousand to two in ten thousand, or from one in ten to two in ten. The clinical response differs. State the observation period, baseline risk, uncertainty, and severity of the outcome. When time changes exposure or risk, use a design and analysis that preserve the temporal sequence.

Confounding by indication is common when a medicine is prescribed because a patient is already at higher risk of the outcome. New-user designs, active comparators, time-aligned exposure definitions, negative controls, and sensitivity analyses may improve inference, but uncertainty should remain visible. The most useful result is one that supports a proportionate clinical decision. Readers should see the numerator, denominator, time window, comparison group, estimate, confidence interval, and limitations before acting on a risk statement.

References

1. PharmacoLearn, Pharm D Year 5 syllabus: https://pharmacolearn.fyi/syllabus/pharm-d/year-5/
2. Pharmacy Council of India, Pharm.D. Regulations, 2008.
3. Rothman, Greenland and Lash, Modern Epidemiology, latest available edition.

Unit 4: Pharmacoepidemiological Study Methods

Syllabus scope

This unit covers drug-utilization review, case reports, case series, surveys, cross-sectional, cohort, case-control, case-cohort, meta-analysis, spontaneous reporting, prescription-event monitoring, and record linkage.

Design selection

Case reports and case series identify unusual events but have no comparison group. Surveys describe self-reported use or attitudes and may be cross-sectional. Cohort studies follow exposed and unexposed groups. Case-control studies begin with an outcome and examine prior exposure. Case-cohort designs sample a subcohort and cases from a defined cohort. Meta-analysis combines studies using explicit methods and must address heterogeneity and bias.

Spontaneous reporting detects signals but is affected by under-reporting, stimulated reporting, missing denominators, and reporting competition. Prescription-event monitoring follows users and solicits events. Record linkage connects prescribing, dispensing, outcomes, and other datasets when identity and privacy controls permit.

Method-selection sequence

1. Define whether the question concerns use, rare harm, incidence, prognosis, or comparative effect.
2. Specify exposure window and outcome window.
3. Choose design that provides a valid comparison when needed.
4. Plan confounder measurement, data quality, and time alignment.
5. Predefine analysis and sensitivity checks.

Clinical correlation

A sudden rise in reports may reflect publicity rather than a true increase. The pharmacist should review exposure volume, reporting completeness, alternative causes, and corroborating data.

Worked example

To study a rare adverse effect, a case-control design may be efficient. Cases and controls need clear definitions, and exposure measurement should use the same source and time rules for both groups.

FAQs

Can a case series estimate incidence? No, not without a valid denominator.

Does meta-analysis guarantee truth? No. It inherits limitations of included studies and synthesis methods.

Takeaways

• Choose design from the question and event frequency.
• Align exposure, outcome, and time definitions.
• Treat signals as prompts for evaluation, not final proof.

Integrated practice note

Every design has a characteristic failure mode. Case reports lack denominators, surveys may suffer recall and nonresponse bias, cohorts can be confounded by treatment selection, case-control studies can mismeasure past exposure, spontaneous reports lack reliable exposure counts, and meta-analyses can combine clinically different studies. A strong protocol names these risks and includes strategies to reduce or test them before interpreting the result.

Study methods should define how participants enter and leave observation, how exposure changes are handled, and how competing events are treated. A patient who dies, switches therapy, or leaves the database may no longer contribute risk in the same way. Time-varying definitions should be planned rather than reconstructed after seeing results. Transparent design makes later replication and critique possible. The protocol should preserve the planned time sequence.

References

1. PharmacoLearn, Pharm D Year 5 syllabus: https://pharmacolearn.fyi/syllabus/pharm-d/year-5/
2. Pharmacy Council of India, Pharm.D. Regulations, 2008.
3. Strom, Kimmel and Hennessy, Textbook of Pharmacoepidemiology, latest available edition.

Unit 5: Data Sources and Special Applications

Syllabus scope

This unit covers ad hoc and automated data sources, vaccine safety, hospital pharmacoepidemiology, risk management, and drug-induced birth defects.

Data sources

Ad hoc sources include chart review, interviews, surveys, registries, and targeted data collection. Automated sources include dispensing records, electronic health records, claims, laboratory systems, registries, and linked databases. Each differs in completeness, coding, access, validation, confounding, and representativeness.

Source-assessment sequence

1. Define the required exposure, outcome, time, and covariates.
2. Check whether the source captures each variable accurately.
3. Validate coding and linkage.
4. Assess missingness, selection, duplication, and follow-up.
5. Protect privacy and document governance.

Vaccine-safety studies must distinguish background disease, coincidental events, expected reactogenicity, and rare signals. Hospital pharmacoepidemiology can examine prescribing, administration, outcomes, antimicrobial use, medication errors, and transitions. Risk management combines detection, evaluation, minimization, communication, and effectiveness assessment.

Pregnancy and birth-defect studies require careful timing by gestational window, exposure measurement, outcome ascertainment, confounding by indication, and comparison selection. A signal needs specialist evaluation and should not be interpreted from a single case alone.

Clinical correlation

A missing prescription record does not prove no exposure; a coded diagnosis does not prove clinical confirmation. Source limitations must be visible in the conclusion.

Worked example

For a vaccine-safety question, a linked registry may provide denominator and timing data, while spontaneous reports may identify an early signal. Combining sources can improve detection but does not remove confounding.

FAQs

Are automated data always better? No. They are large and efficient but may contain coding and completeness problems.

Why is pregnancy timing critical? Risk can depend on developmental stage and exposure window.

Takeaways

• Match source to the variables and validity needed.
• Use multiple evidence streams for safety signals.
• Treat pregnancy and vaccine analyses as time- and context-sensitive.

Integrated practice note

Data linkage can increase power while increasing privacy and misclassification risk. Use the minimum necessary identifiers, document linkage rules, assess false matches and missed matches, restrict access, and retain an audit trail. Special applications need careful clinical review because a coded record may not capture severity, clinician reasoning, patient preference, or the exact exposure that occurred.

Validation can include chart review, laboratory confirmation, clinician assessment, or comparison with an independent registry. The level of validation should match the consequence of misclassification. Privacy protections should not prevent legitimate safety action, but safety action should still use authorized pathways and minimum necessary data. Data governance is part of scientific validity and participant protection.

Special applications should predefine clinical review. Vaccine-safety assessment needs a denominator and background rate. Hospital studies need a clear admission, exposure, and discharge window. Birth-defect studies need gestational timing and appropriate comparison. Risk-management studies need an effectiveness measure. These details prevent a broad label from concealing the specific bias or uncertainty that determines interpretation.

References

1. PharmacoLearn, Pharm D Year 5 syllabus: https://pharmacolearn.fyi/syllabus/pharm-d/year-5/
2. Pharmacy Council of India, Pharm.D. Regulations, 2008.
3. World Health Organization pharmacovigilance and vaccine-safety resources.

Unit 6: Pharmacoeconomic Foundations and Outcomes

Syllabus scope

This unit covers the definition, history, need, and use of pharmacoeconomics, its role in formulary decisions, outcome assessment, and types of evaluation.

Pharmacoeconomics compares costs and consequences of alternatives. It supports decisions about value, affordability, access, and opportunity cost. It does not mean choosing the cheapest medicine. The analysis should identify perspective, time horizon, alternatives, costs, outcomes, discounting where relevant, uncertainty, and distributional concerns.

Evaluation pathway

1. Define the decision and competing alternatives.
2. Choose perspective: patient, provider, payer, or society.
3. Identify resource use and outcomes.
4. Measure and value costs and effects.
5. Calculate incremental results.
6. Test assumptions and uncertainty.
7. Interpret affordability, equity, and clinical relevance.

Costs may include medicine, administration, monitoring, adverse events, hospitalization, transport, productivity, and patient time. Outcomes may be natural units, monetary benefits, quality-adjusted life years, life-years, symptoms, or events avoided.

Formulary role

Formulary committees use economic evidence with efficacy, safety, quality, supply, patient preference, and budget impact. A favorable cost-effectiveness result does not automatically mean the budget can afford immediate adoption.

Worked example

If a new product costs more but prevents admissions, the committee should compare incremental medicine and care costs with incremental outcomes over a defined time horizon. The conclusion depends on perspective and assumptions, not acquisition price alone.

FAQs

Is pharmacoeconomics the same as budget impact? No. Cost-effectiveness asks value per outcome; budget impact asks affordability for a defined budget and population.

Who chooses the perspective? It should be stated and justified for the decision.

Takeaways

• Compare alternatives, costs, outcomes, and uncertainty.
• State perspective and time horizon.
• Combine value evidence with access and budget realities.

Integrated practice note

Economic evaluation should not hide ethical choices. A model may favor an intervention with a high average benefit while leaving people with complex needs behind. Report distribution, equity, affordability, and implementation constraints alongside the ratio. The pharmacist can help ensure that resource-use assumptions reflect actual preparation, monitoring, adherence support, adverse-event management, and patient access rather than only acquisition price.

Perspective changes the answer. A payer may count medicine and hospital costs, a patient may count travel, time, and out-of-pocket payment, and society may include productivity. State whose costs and outcomes are included instead of blending them into an unexplained total. Economic evidence should complement, not replace, clinical judgement. Document assumptions and their sources clearly.

Outcome assessment in practice

Outcome selection should reflect what the intervention is expected to change. A medicine may lower a laboratory value without improving symptoms, function, admissions, or survival. A service intervention may change adherence or access before a clinical endpoint changes. Define the causal pathway and choose outcomes at the relevant level. Report both intended benefits and harms, including treatment burden, monitoring burden, discontinuation, and patient-reported experience.

When outcomes differ in importance, present them separately before combining them in a utility or composite measure. Explain the source of utility values and whether they reflect the studied population. If quality-of-life data are missing, do not fill the gap with an invented value. Treat the input as uncertain, perform sensitivity analysis, and show how much the conclusion depends on it.

References

1. PharmacoLearn, Pharm D Year 5 syllabus: https://pharmacolearn.fyi/syllabus/pharm-d/year-5/
2. Pharmacy Council of India, Pharm.D. Regulations, 2008.
3. Drummond et al., Methods for the Economic Evaluation of Health Care Programmes, latest available edition.

Unit 7: Pharmacoeconomic Evaluation Methods

Syllabus scope

This unit covers cost-minimization, cost-benefit, cost-effectiveness, and cost-utility analysis.

Methods

Cost-minimization analysis compares costs only when outcomes are demonstrated equivalent. Cost-benefit analysis converts both costs and benefits to monetary units. Cost-effectiveness analysis compares cost with a natural clinical outcome, such as cases prevented or life-years gained. Cost-utility analysis commonly uses quality-adjusted life years, combining survival and health-related quality.

Analysis sequence

1. Confirm the alternatives are clinically relevant.
2. Choose perspective and time horizon.
3. Identify all relevant cost and outcome components.
4. Measure and value each component consistently.
5. Calculate incremental cost and incremental effect.
6. Present ratios only with context and uncertainty.
7. Conduct sensitivity analysis.

The incremental cost-effectiveness ratio is incremental cost divided by incremental effect. It is not meaningful when the new option is both more costly and less effective without a compelling nonclinical rationale. Dominance and extended dominance should be assessed.

Worked example

If option A costs 100 units and produces 10 outcome units, while B costs 130 and produces 12, the incremental cost is 30 and incremental effect is 2, giving 15 cost units per additional outcome unit. Whether that is acceptable depends on the decision context and threshold, which must not be invented.

Quality and uncertainty

Quality assessment examines perspective, comparators, discounting, modeling, data sources, conflicts, sensitivity analysis, and transferability. Probabilistic sensitivity analysis can show how results change across uncertain inputs.

FAQs

When is cost-minimization valid? Only after outcome equivalence is credible.

Is a cost-effectiveness ratio a price? No. It is an incremental comparison linked to a defined outcome.

Takeaways

• Choose method from outcome equivalence and decision question.
• Report incremental costs and effects.
• Test uncertainty and avoid unsupported thresholds.

Integrated practice note

Cost-minimization is particularly vulnerable to an unsupported assumption of equal outcomes. Cost-benefit analysis can be difficult when benefits are uncomfortable to value in money. Cost-effectiveness depends on a meaningful natural outcome, while cost-utility depends on utility measurement and quality-of-life assumptions. Whichever method is used, disclose the comparator, perspective, time horizon, costing source, outcome definition, discounting, and sensitivity analysis.

Discounting, inflation, currency year, and transferability matter for longer analyses. Use a consistent costing year and disclose conversion methods. Sensitivity analysis should change uncertain inputs across plausible ranges and show whether the decision changes. If assumptions dominate the result, decision makers need that warning clearly.

References

1. PharmacoLearn, Pharm D Year 5 syllabus: https://pharmacolearn.fyi/syllabus/pharm-d/year-5/
2. Pharmacy Council of India, Pharm.D. Regulations, 2008.
3. Drummond et al., Methods for the Economic Evaluation of Health Care Programmes, latest available edition.

Unit 8: Applications, Software and Case Studies

Syllabus scope

This unit covers applications of pharmacoeconomics, software, and case studies.

Application workflow

Applications include formulary selection, antimicrobial stewardship, chronic-disease programs, screening, vaccination, hospital service design, adherence support, and medicine-safety interventions. A case study should state the decision, alternatives, population, perspective, data, model, assumptions, result, uncertainty, and implementation implication.

Case-study sequence

1. Frame the decision and stakeholders.
2. Identify clinical evidence and economic inputs.
3. Build a transparent calculation or model.
4. Test alternative assumptions.
5. Present results in understandable tables and diagrams.
6. Decide how the evidence will affect practice.
7. Monitor real-world outcomes after implementation.

Software can support data cleaning, epidemiologic analysis, decision trees, Markov models, cost calculations, sensitivity analysis, and visualization. It cannot choose a valid perspective, fix biased inputs, or decide whether an outcome is clinically meaningful. Check the currently supported software version and institutional tools before beginning an analysis.

Clinical correlation

An economic model can appear precise while depending on uncertain transition probabilities, adherence, utility values, costs, or treatment effects. Decision makers need ranges, scenario analysis, and an explanation of what would change the conclusion.

Worked example

A hospital compares two discharge-support programs. The analysis includes staff time, patient contacts, readmissions, adherence, quality of life, and implementation capacity. The cheaper program is not automatically preferred if it produces worse outcomes or cannot reach the intended patients.

FAQs

Can software replace economic reasoning? No. It performs specified calculations and reflects the assumptions entered.

Why monitor after implementation? Real-world utilization and outcomes may differ from the model.

Takeaways

• Make case-study assumptions visible.
• Use software reproducibly and test uncertainty.
• Link economic results to implementation and outcomes.

Integrated practice note

Implementation is the final test of an economic recommendation. Monitor whether the expected population receives the intervention, whether resource use matches assumptions, whether adherence and outcomes change, and whether access becomes unequal. Update the analysis when prices, guidelines, resistance, staffing, or clinical outcomes change. A model should support a decision, not make the decision invisible.

Software output should be checked against hand calculations or simple test cases. Preserve input files, formulas, model version, assumptions, and output date. A reproducible case study lets another reviewer identify whether disagreement comes from data, method, value judgement, or implementation constraint. The final report should state who can act on the result and how it will be reviewed. Review findings periodically.

Case-study interpretation framework

Begin with the decision, not the software. State which patients are being considered, what choices are available, what outcomes matter, and which perspective is used. Then identify the evidence supporting clinical effectiveness, utilization, adverse events, adherence, and costs. If the case uses a decision tree, define each branch and probability. If it uses a Markov model, define health states, transition probabilities, cycle length, time horizon, and stopping or absorbing states. Check whether patients can move between states in a clinically plausible way.

Next, separate observed inputs from assumptions. Observed costs may come from local procurement or accounting data, while treatment effects may come from a trial and utilities from a published survey. Each source has a population and time context. Convert currency and price year transparently, avoid double counting, and state whether costs are discounted. Present base-case results, one-way sensitivity analysis, scenario analysis, and probabilistic analysis when appropriate.

Finally, explain implementation. A program that is cost-effective may still require staff, training, equipment, supply continuity, patient travel, or monitoring that the model did not include. Review uptake, outcomes, adverse events, equity, and actual expenditure after implementation. Update the model when the clinical pathway or price changes. This makes pharmacoeconomics a practical decision tool rather than a single ratio detached from care.

References

1. PharmacoLearn, Pharm D Year 5 syllabus: https://pharmacolearn.fyi/syllabus/pharm-d/year-5/
2. Pharmacy Council of India, Pharm.D. Regulations, 2008.
3. Drummond et al., Methods for the Economic Evaluation of Health Care Programmes, latest available edition.