Clinical Trials, Biostatistics
What the numbers mean
Phase I establishes dose and toxicity, phase II screens for activity, and phase III compares outcomes against a control. Overall survival is the most robust endpoint; progression-free survival is accepted when it reliably predicts overall survival but is vulnerable to assessment bias and differing definitions. Randomization plus intention-to-treat analysis preserves prognostic balance; per-protocol analysis can introduce bias by removing noncompliant or sicker patients.
The hazard ratio summarizes time-to-event comparisons under proportional hazards, while the log-rank test and Cox models provide inference. Power depends on effect size, alpha, sample size, and event count. A p-value is not the probability the hypothesis is true; it is the probability of data at least this extreme if the null were true. Confidence intervals convey both effect size and precision. Noninferiority trials require a prespecified margin and are invalid if the control arm underperforms or if assay sensitivity is absent. Interim analyses spend alpha, and multiple comparisons require correction. Subgroup findings should be hypothesis-generating unless prespecified and powered.
Traps
Screening trials can suffer lead-time bias; surrogate endpoints can overstate benefit; poor accrual or crossover complicates interpretation; an unplanned subgroup is not a practice changer.
Worked example
A trial reports a hazard ratio of 0.75 with 95% CI 0.60–0.94. Because the interval excludes 1.0 and the upper bound lies below 1.0, the result is statistically significant at alpha 0.05 and suggests a 25% relative reduction in the hazard of the event.
Supportive and Palliative Care
Symptom control is a core oncologic skill
Pain management follows the analgesic ladder, with opioids titrated to effect, always paired with a bowel regimen because constipation is predictable. In renal impairment, morphine accumulates active metabolites; hydromorphone or fentanyl is preferred. Chemotherapy-induced nausea is prevented rather than treated, with a 5-HT3 antagonist, dexamethasone, and an NK1 antagonist for highly emetogenic regimens such as cisplatin, often with olanzapine. Febrile neutropenia requires urgent cultures and empiric broad-spectrum antibiotics; colony-stimulating factors are used for secondary or high-risk primary prophylaxis. Neuropathy, mucositis, and fatigue require dose modification and supportive measures rather than abandonment of curative intent.
Early integrated palliative care improves quality of life and, in some trials, survival, and it is appropriate alongside curative treatment - it is not synonymous with hospice. Goals-of-care discussions should be documented, revisited at transitions, and used to guide code status and site of care. Hospice eligibility generally requires a prognosis of 6 months or less if the disease follows its natural course. Malignant bowel obstruction in ovarian cancer is managed with antisecretory therapy, corticosteroids, and decompression, with surgery reserved for selected patients.
Traps
Equating palliative care with giving up; forgetting bowel prophylaxis; using megestrol for appetite with the expectation of survival benefit; failing to reassess goals after a major complication.
Worked example
A patient receiving cisplatin needs antiemetic prophylaxis before the infusion: a 5-HT3 antagonist, dexamethasone, an NK1 antagonist, and olanzapine for a highly emetogenic regimen, not rescue therapy after vomiting begins.