Using QALYs to express outcomes is what cost-utility analysis does. QALYs combine how long people live with the quality of that life into a single measure, so you can compare different health interventions even if they affect different aspects of health. In cost-utility analysis, you look at costs relative to QALYs gained, yielding metrics like cost per QALY gained that help assess value for money when both longevity and quality matter. Other analyses don’t use QALYs: cost-minimization assumes outcomes are the same and only compares costs; cost-consequence lists multiple outcomes without aggregating them into one summary; cost-benefit translates benefits into monetary terms rather than into quality-adjusted life years.

Costs tend to cluster at lower values with a long tail to higher costs. In real-world cost data, most items are relatively inexpensive, but a few items incur very high costs, pulling the tail to the right. Since costs can’t be negative, that creates a right-skewed distribution, where the mean is pulled upward by the few large costs and the median sits closer to the bulk of the data. This shape matters for analysis: the median often provides a more representative sense of typical cost, and analysts frequently use transformations or models suited for skewed data (for example, log-transformations or gamma-type models). Left-skewed would imply a tail on the cheaper side, which isn’t typical for costs. A bimodal pattern would suggest two distinct cost groups rather than one gradual spread. A normal distribution would be symmetric, which doesn’t match the common skew toward higher-cost outliers.

Direct costs are the actual medical resources used to deliver care. This means the elements of care itself—the treatment provided, the medications used, and the equipment employed during that care. These are the expenses tied directly to administering the medical intervention. Productivity loss represents the economic impact of illness on work and is an indirect cost. Transportation and patient time costs reflect non-medical aspects of care—costs associated with getting to and being in care or the time spent, rather than the medical resources used. So, the combination of treatment, medications, and equipment fits direct medical costs, while the others fall into indirect or non-medical cost categories.

Bootstrapping is a resampling approach used to quantify uncertainty in economic evaluations without assuming a specific parametric distribution for costs and effects. In economic evaluations, cost data are often right-skewed because a few individuals incur very high costs, and QoL measures can be non-normally distributed. This makes traditional parametric methods unreliable or unstable. By repeatedly drawing samples with replacement from the observed data and recalculating the incremental cost and effect in each resample, you build an empirical distribution that reflects the actual variability and the relationship between costs and outcomes. This lets you derive confidence intervals for costs and QALYs and create cost-effectiveness acceptability curves that show the probability an intervention is cost-effective at different willingness-to-pay thresholds. The advantage is robustness to skewness and small sample issues, rather than assuming normality.