Oster shows how absolute risk, confounding, study design, and personal preference can turn a scolding rule into a useful question for a prenatal appointment.
Is Expecting Better worth reading?
An excellent antidote to unexplained pregnancy rules and a useful lesson in reading risk, but not a substitute for prenatal care. Trust its questions more than any medical conclusion that conflicts with current guidance.
Full review
Why it works, and where it does not
What it is
Oster applies an economist’s decision framework to pregnancy: separate evidence from convention, distinguish correlation from causation, compare absolute rather than relative risk, and include the pregnant person’s preferences and costs. This is genuinely empowering when a rule is preference-sensitive or based on weak evidence. The limitation is equally important: population studies cannot diagnose an individual pregnancy, absence of convincing evidence of harm is not proof of safety, and public-health guidance may rationally favor a precaution when an exposure has no medical benefit and fetal harm cannot be ruled out.
The confident economic frame can make uncertain observational evidence feel more decisive than it is, and advice can date quickly as screening options and medical guidance change.
Discusses miscarriage, fetal abnormality, prenatal testing, alcohol and caffeine exposure, foodborne illness, labor decisions, and pregnancy complications. It cannot personalize medication or high-risk-pregnancy decisions, and current clinical guidance should take priority.
The honest critique
The book is strongest when it exposes vague rules and asks for denominators, confounders, and actual effect sizes. It is weaker when its own reading of observational evidence becomes a recommendation. Alcohol is the clearest example: studies of light exposure are inconsistent and difficult to interpret because drinking is misreported and correlated with income, diet, smoking, and other factors. That uncertainty does not establish a safe dose. CDC and ACOG continue to advise that no amount, timing, or type of alcohol is known to be safe in pregnancy. Caffeine is different: current ACOG guidance says intake below 200 mg per day does not appear to be a major contributor to miscarriage or preterm birth, broadly supporting skepticism toward blanket coffee bans. Oster’s miscarriage-by-week discussion can reduce misplaced self-blame, but population averages vary with age, history, symptoms, ultrasound findings, and other clinical details; they cannot determine whether one pregnancy is viable or whether bleeding needs assessment. Prenatal testing is especially vulnerable to edition age. Current ACOG guidance says screening and diagnostic options should be discussed and offered to all pregnant patients regardless of age or baseline risk, and a positive cell-free DNA screen requires counseling and an offer of confirmatory CVS or amniocentesis because screening is not diagnosis.
Choose this instead when
Choose the newest Mayo Clinic Guide to a Healthy Pregnancy if you want a broader clinical reference organized around pregnancy stages, symptoms, and care rather than an economist’s argument about selected rules.
Source check
Sources behind this review
Show all 7 sourcesHide additional sources
Put it to work
Try the useful part in real life.
Write the decision, the current guideline, the absolute benefit and harm ranges, what is uncertain, and the personal fact that may change your baseline risk. Bring the sheet to your prenatal clinician. Do not use a population average to override advice based on your conditions, medications, history, or current symptoms.
Ask whether the choice has a medical benefit, whether a safe threshold is established, and how serious an uncertain downside could be. For caffeine, keep total intake within the current clinician-agreed limit, commonly below 200 mg daily under ACOG guidance. For alcohol, follow current CDC and ACOG abstinence guidance rather than treating inconclusive low-dose studies as proof of safety.
Ask what a test screens for, its false-positive and false-negative limits in your risk group, whether the result is diagnostic, and what you would do next. A high-risk cell-free DNA result should lead to genetic counseling and discussion of diagnostic confirmation, not an irreversible decision based on screening alone. Seek care for concerning symptoms regardless of the book’s miscarriage probabilities.
Questions to make you think
- Is this conclusion evidence of no harm, or merely no clear evidence of harm in studies too confounded or underpowered to settle the question?
- What is my absolute baseline risk, and which facts about my pregnancy make the book’s population average a poor fit?
- Has the relevant guideline, test technology, or edition changed since the evidence Oster reviewed?
- Which choice is genuinely preference-sensitive, and which one has a current clinical safety recommendation for reasons the book may underweight?
Did this review help you decide?
Your answer is saved and helps us decide what to revise next.