PGIM MD Community Medicine common mistakes
PGIM MD Community Medicine · Understand the traps and how to avoid them.
Writing descriptive epidemiological answers as if they were clinical case reports, describing individual patients and their treatment rather than rates, denominators, and population-level comparisons.
Most candidates train in hospital settings where the unit of analysis is the patient; community medicine questions explicitly ask for population measures, and habitual clinical framing loses marks even when the underlying knowledge is correct.
What to do instead
Before writing, force yourself to state the numerator and denominator and the population at risk. Practise converting any clinical scenario into incidence, prevalence, and attributable risk statements, and make the first sentence of every answer a population-level statement.
Mixing up incidence and prevalence, and using them interchangeably when asked to interpret a screening or follow-up dataset.
Both are expressed as proportions or rates per population and share the word 'rate' in casual speech; under time pressure candidates grab whichever term comes first instead of reasoning about new cases versus existing cases.
What to do instead
Anchor on the definition: incidence counts new onsets in a defined period in a population at risk; prevalence counts all existing cases at a point or over a period. For every practice question, write the definition in the margin before choosing, and note that prevalence approximates incidence times duration for stable diseases.
Confusing the measures of association and effect: reporting a relative risk where an odds ratio was required, or interpreting a risk difference as a ratio.
Case-control and cohort designs are taught together, and candidates memorise formulas without linking each formula to its study design and its interpretation, so they compute the wrong quantity confidently.
What to do instead
Build a one-page table mapping design to measure: cohort gives risk ratio and risk difference; case-control gives odds ratio only; cross-sectional gives prevalence ratio. For each, write one sentence of interpretation, then practise identifying the design from a short vignette before computing anything.
Declaring a result statistically significant without checking whether confounding, bias, or chance could explain it, and ignoring confidence intervals in favour of the p-value alone.
Exams and clinical teaching emphasise p-values, so candidates treat a small p as proof of a real effect and skip the appraisal steps that community medicine marks reward.
What to do instead
Use a fixed appraisal sequence for every study: design, selection of participants, measurement of exposure and outcome, confounding control, then precision via confidence intervals. State explicitly whether the interval crosses the null and whether the point estimate is large enough to matter, before mentioning significance.
Confusing sensitivity and specificity with positive and negative predictive values, especially when asked what a test result means for a patient.
All four are conditional probabilities and the formulas look similar; candidates also forget that predictive values depend on prevalence while sensitivity and specificity do not.
What to do instead
Draw a two-by-two table for every screening question and fill in numbers rather than manipulating symbols. Remember the mnemonic that sensitivity and specificity are properties of the test, while predictive values are properties of the test in a population; test yourself by changing prevalence and predicting which values move.
Answering outbreak investigation questions by jumping straight to control measures without describing case definition, line listing, and hypothesis generation in sequence.
Candidates remember the dramatic interventions and forget that the marks are distributed across the standard steps of an investigation, which are examined in order.
What to do instead
Memorise and rehearse the sequence: verify the diagnosis and confirm the outbreak, establish a case definition, identify and count cases, describe by person, place, and time, generate hypotheses, test them with an analytical study, then implement and evaluate control. Write the steps as headings and place your content under each.
Describing national health programmes from memory with outdated names, delivery structures, or eligibility criteria, or inventing a plausible-sounding scheme.
Health programmes are renamed, merged, and restructured frequently, and candidates rely on notes from earlier training or on general impressions rather than current official documents.
What to do instead
Keep a dated one-page summary per programme covering objectives, target group, package of services, and level of delivery, and refresh it from the current official programme guidelines before the exam. If you cannot recall a specific detail with confidence, describe the programme's function and structure generically instead of inventing names or figures.
Treating maternal and child health or nutrition questions as pure clinical management, omitting the community-level components such as growth monitoring, immunisation scheduling, and referral pathways.
Clinical training rewards management decisions, so candidates write about drugs and dosages when the question asks about service delivery, coverage, and follow-up in a defined population.
What to do instead
For every maternal and child health answer, deliberately cover the continuum of care: antenatal, intrapartum, postnatal, newborn, and child health contacts, plus the platform that delivers them. Add coverage indicators and referral triggers, and check that at least half your answer addresses services and populations rather than individual treatment.
Presenting health planning and management answers as unstructured lists of activities, without objectives, indicators, resources, monitoring, or evaluation.
Planning frameworks are abstract and candidates reduce them to remembered buzzwords, which reads as a list rather than a plan and loses marks for logical structure.
What to do instead
Use a fixed planning skeleton every time: situational analysis, priority setting with explicit criteria, objectives that are specific and measurable, strategies and activities, resource requirements, indicators, and monitoring and evaluation. Practise writing one short plan under exam timing so the skeleton becomes automatic.
Misapplying biostatistics by choosing a test or summary measure that does not match the data type and distribution, for example using a mean and standard deviation for skewed data or a parametric test for small non-normal samples.
Candidates memorise test names and their conditions separately from the data description, so the decision about which test to use is made after the data have already been summarised inappropriately.
What to do instead
Work in a fixed order: identify the variable type, check the distribution and sample size, choose the summary measure, then choose the test. Keep a decision table linking outcome type and comparison to the correct test, and for skewed data default to median and interquartile range with a non-parametric test.