Obstetrics and Gynaecology chapters
Maternal and perinatal statistics for Pakistan and study design
How maternal and perinatal deaths are defined and measured, what national surveys show for Pakistan, why mothers die, and the study designs, 2x2 tables and risk measures that obstetric research and MCQs rely on.
- A maternal death is a death in pregnancy or within 42 days of its end from a cause linked to or worsened by the pregnancy (not accidents); the maternal mortality ratio is deaths per 100,000 live births.
- Perinatal mortality counts stillbirths plus early neonatal deaths (first 7 days) per 1,000 total births; in Pakistan the national survey of 2019 gave an MMR of 186, with haemorrhage the leading cause.
- Case-control studies give an odds ratio, cohort studies give relative risk and incidence, and the randomised controlled study is the best single design for testing a treatment; a 2x2 table gives sensitivity, specificity and predictive values.
On this page
Overview
Maternal and perinatal deaths are the clearest measures of how well a country cares for mothers and babies. PGET questions test the exact definitions (time windows and denominators), the main causes of maternal death in Pakistan, and the basic research methods used to produce obstetric evidence.
Most maternal deaths worldwide are preventable, and almost all of them occur in low- and middle-income countries. Ten Teachers notes that around 830 women died each day in 2016 from pregnancy-related causes, 99% of them in low-income settings.
Definitions
Maternal deaths
| Term | Definition |
|---|---|
| Maternal death | Death of a woman while pregnant or within 42 days of the end of pregnancy, whatever the length or site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management; accidental or incidental causes are excluded |
| Late maternal death | Death from such causes after 42 days and up to 1 year after the end of pregnancy |
| Direct | From a complication of pregnancy itself: haemorrhage, hypertensive disorders, sepsis, abortion complications, VTE, amniotic fluid embolism |
| Indirect | From a pre-existing or new disease made worse by pregnancy: heart disease, epilepsy, anaemia, hepatitis, influenza and other infections |
| Coincidental (incidental) | Unrelated causes such as road accidents; not counted in the MMR |
Measures
| Measure | Numerator | Denominator |
|---|---|---|
| Maternal mortality ratio (MMR) | Maternal deaths | Per 100,000 live births |
| Maternal mortality rate | Maternal deaths | Per 100,000 women of reproductive age (15-49) |
| Perinatal mortality rate (PMR) | Stillbirths + early neonatal deaths (first 7 days) | Per 1,000 total births (live births + stillbirths) |
| Stillbirth rate | Babies born dead after the viability threshold | Per 1,000 total births |
| Early neonatal mortality | Deaths in the first 7 days | Per 1,000 live births |
| Neonatal mortality rate | Deaths in the first 28 days | Per 1,000 live births |
| Infant mortality rate | Deaths under 1 year | Per 1,000 live births |
- Stillbirth threshold: WHO uses 28 weeks (or 1,000 g) for international comparison; the UK uses 24 weeks. Losses before the threshold are miscarriages.
- The MMR uses live births as the denominator because the true number of pregnancies is hard to count.
The MMR is per 100,000 live births, while neonatal, infant and perinatal rates are per 1,000. The PMR uses total births (live + still), not live births alone.
A district records 30 maternal deaths and 15,000 live births in a year
MMR = 30 ÷ 15,000 × 100,000 = 200 per 100,000 live births
980 live births + 20 stillbirths = 1,000 total births
10 babies die in the first week
PMR = (20 + 10) ÷ 1,000 × 1,000 = 30 per 1,000 total births
"Peri = around birth": just before (stillbirth) and one week after (early neonatal death).
Pakistan: key figures
These figures come from national surveys run by the National Institute of Population Studies (NIPS) under the Ministry of National Health Services. Newer surveys may change them; quote the latest official report.
| Indicator | Figure | Source |
|---|---|---|
| Maternal mortality ratio | 186 per 100,000 live births (3 years before the survey; down from 276 in 2006-07) | Pakistan Maternal Mortality Survey 2019 |
| Highest provincial MMR | Balochistan (298) | PMMS 2019 |
| Leading cause of maternal death | Obstetric haemorrhage, about 41% (up from 33% in 2006-07) | PMMS 2019 |
| Abortion-related deaths | About 10% | PMMS 2019 |
| Direct vs indirect | About 96% direct, 4% indirect | PMMS 2019 |
| Infant mortality | 62 per 1,000 live births | PDHS 2017-18 |
| Under-5 mortality | 74 per 1,000 live births | PDHS 2017-18 |
| Skilled antenatal care at least once | 86% | PDHS 2017-18 |
| Delivery in a health facility | 66% | PDHS 2017-18 |
| Total fertility rate | 3.6 children per woman | PDHS 2017-18 |
| Contraceptive use (married women) | 34% | PDHS 2017-18 |
Official sources: NIPS: Pakistan Maternal Mortality Survey 2019, policy brief (checked 11 Nov 2026); NIPS: PMMS 2019 policy brief with cause breakdown (checked 11 Nov 2026); Government of Pakistan PID: key findings of PDHS 2017-18 (checked 11 Nov 2026); WHO: maternal mortality ratio definition (checked 11 Nov 2026).
In Pakistan, the main killers are direct obstetric causes: haemorrhage first, then hypertensive disorders (pre-eclampsia and eclampsia), then abortion complications and pregnancy-related infection. Nearly all are preventable with skilled birth care and timely referral.
In high-income countries such as the UK, indirect causes (heart disease, VTE, sepsis, mental health) now dominate, and VTE has been the leading direct cause. Do not transfer UK cause rankings to Pakistan.
Why mothers die: the three delays
| Delay | Examples | Fixes |
|---|---|---|
| 1. Deciding to seek care | Not recognising danger signs, cost, family decision-making, low status of women | Health education, birth preparedness, community health workers |
| 2. Reaching care | Distance, transport, roads, cost of travel | Ambulance and referral systems, nearby facilities |
| 3. Receiving adequate care | Shortage of staff, blood, drugs, theatre; poor skills | Emergency obstetric care, blood banks, training, audit |
Other contributors: anaemia, high parity, short birth spacing, early marriage, home births without a skilled attendant, unsafe abortion and poor access to contraception.
Interventions that save mothers: skilled birth attendance, access to emergency obstetric care (including caesarean and blood), active management of the third stage, magnesium sulfate for eclampsia, clean delivery and antibiotics, family planning, antenatal care, tetanus toxoid immunisation, iron and folate.
Audit: confidential enquiries into maternal deaths (such as the UK MBRRACE reports) and maternal death surveillance and response reviews each death to find avoidable factors.
Study design
Types of study
| Design | What it does | Direction | Main measure | Strength | Weakness |
|---|---|---|---|---|---|
| Case report / case series | Describes unusual patients | - | - | Spots new diseases | No comparison group |
| Cross-sectional | Measures exposure and disease at one point in time | Snapshot | Prevalence | Quick, cheap; good for surveys (e.g. PDHS) | Cannot show which came first |
| Ecological | Compares groups, not individuals | - | Correlation | Uses routine data | Ecological fallacy |
| Case-control | Starts with disease (cases) and without (controls), looks back at exposure | Retrospective | Odds ratio | Best for rare diseases (e.g. congenital anomalies); quick | Recall and selection bias |
| Cohort | Starts with exposure, follows for outcome | Usually prospective | Incidence, relative risk, attributable risk | Shows time sequence; several outcomes | Slow, costly; poor for rare outcomes; loss to follow-up |
| Randomised controlled study (RCT) | Randomly allocates an intervention | Prospective | Relative risk reduction, NNT | Best design to test treatment; randomisation balances confounders | Costly; ethical limits in pregnancy |
| Systematic review and meta-analysis | Pools all good studies on a question | - | Pooled estimate (forest plot) | Top of the evidence hierarchy | Only as good as the included studies |
Hierarchy of evidence (highest first)
- Systematic review and meta-analysis of RCTs
- Randomised controlled study
- Cohort study
- Case-control study
- Cross-sectional study, case series and case reports
- Expert opinion
Key terms
- Randomisation: each participant has an equal chance of each group; removes selection bias and balances known and unknown confounders.
- Blinding: single (participant), double (participant and investigator), triple (also the analyst). Reduces observer and reporting bias.
- Confounding: a third factor linked to both exposure and outcome (for example maternal age when studying parity and Down syndrome). Controlled by randomisation, matching, restriction or statistical adjustment.
- Intention-to-treat analysis: analyse everyone in the group they were randomised to.
- Incidence = new cases in a period; prevalence = all existing cases at a time point. Prevalence ≈ incidence × average duration.
- P value below 0.05 is usually called statistically significant; a 95% confidence interval that crosses 1 (for a ratio) means no significant difference.
Case-control → odds ratio; cohort → relative risk. For a rare outcome, choose a case-control study.
Measures of risk
Cohort: GDM in 60 of 500 obese women (12%) and 20 of 500 non-obese women (4%)
Relative risk = 12% ÷ 4% = 3; attributable risk = 12% − 4% = 8%
Case-control: 40 of 100 stillbirth cases and 20 of 100 controls smoked
Odds ratio = (40 × 80) ÷ (60 × 20) = 3,200 ÷ 1,200 = 2.67
A drug lowers the risk of an outcome from 4% to 2%
Absolute risk reduction = 4% − 2% = 2% = 0.02
NNT = 1 ÷ 0.02 = 50
Ten Teachers gives a real example: with risk-based GBS prophylaxis, about 5,882 women need treatment to prevent one neonatal death, showing how a large NNT arises when the outcome is rare.
Screening and diagnostic tests
A good screening test must target an important condition, change management, and have high sensitivity, specificity and predictive value, while being affordable and equally available to all.
| Test result | Disease present | Disease absent |
|---|---|---|
| Positive | True positive (a) | False positive (b) |
| Negative | False negative (c) | True negative (d) |
- Sensitivity = a ÷ (a + c): proportion of the sick correctly detected.
- Specificity = d ÷ (b + d): proportion of the healthy correctly cleared.
- Positive predictive value (PPV) = a ÷ (a + b): chance that a positive result is a true case; rises with prevalence.
- Negative predictive value (NPV) = d ÷ (c + d).
TP 90, FN 10, FP 30, TN 170
Sensitivity = 90 ÷ 100 = 90%
Specificity = 170 ÷ 200 = 85%
PPV = 90 ÷ 120 = 75%
NPV = 170 ÷ 180 = 94.4%
"SnNout, SpPin": a Sensitive test, when Negative, rules out; a Specific test, when Positive, rules in.
Expect: the 42-day window for maternal death and late deaths up to 1 year; MMR per 100,000 live births versus neonatal, infant and perinatal rates per 1,000; perinatal mortality as stillbirths plus first-week deaths over total births; direct versus indirect versus coincidental deaths; Pakistan's MMR of 186 (PMMS 2019), the highest figure in Balochistan, and haemorrhage as the leading cause followed by hypertensive disorders; the three delays; matching study designs to measures (case-control with odds ratio, cohort with relative risk, cross-sectional with prevalence); RCT and meta-analysis at the top of the evidence hierarchy; calculations of sensitivity, specificity, predictive values, odds ratio and NNT.
Quick revision
- Maternal death: in pregnancy or within 42 days, related or aggravated, not accidental; late = 42 days to 1 year.
- MMR per 100,000 live births; neonatal (28 days) and infant (1 year) per 1,000 live births.
- PMR = stillbirths + early neonatal deaths per 1,000 total births.
- Pakistan MMR 186 (PMMS 2019); Balochistan highest; haemorrhage the leading cause; 96% direct.
- Three delays: deciding, reaching, receiving care.
- Case-control: rare disease, odds ratio. Cohort: incidence, relative risk. RCT: best for treatment.
- Sensitivity a/(a+c); specificity d/(b+d); PPV a/(a+b), rises with prevalence.
Written by The House Job from standard references (Obstetrics by Ten Teachers), with national figures from NIPS and Government of Pakistan reports and definitions from WHO.
FAQ
Why does the maternal mortality ratio use live births and not pregnancies?
Live births are recorded far more reliably than all pregnancies, many of which end early without being counted, so live births give a practical, comparable denominator.
Is a death from a road accident in pregnancy a maternal death?
No. Accidental and incidental causes are excluded; it is recorded as a coincidental death and does not enter the maternal mortality ratio.
Which study design would you use to find whether a drug taken in early pregnancy causes a rare birth defect?
A case-control study, starting with babies who have the defect and comparable babies without it, then comparing past drug exposure; the result is an odds ratio.
Why does the positive predictive value of a screening test fall in a low-risk population?
When a condition is rare, even a small false-positive rate produces many false positives compared with true cases, so a positive result is less likely to be genuine.