--- module: 103-03 language: en chapter: 103 title: "Social Determinants, Health Systems, Economics, and Global Burden" module_title: "Disease burden, implementation, global health, and planetary health" source_sha256: e09f3bcd61206e101f4268b7c0bdf9dd58f2d5de24c4e7f3c931eb33054d0c19 --- # Disease burden, implementation, global and planetary health ## Measuring burden ### Counts depend on population size ### Rates need numerator, denominator, time ### Incidence: new events among those at risk ### Prevalence: existing disease #### Rises with incidence and duration #### Better survival can raise prevalence ### Case fatality depends on detection and definition ### Population mortality also depends on incidence ### Death certification carries uncertainty ## Standardisation and summary measures ### Crude rates mislead across age structures ### Direct: rates applied to a standard population ### Indirect: observed versus expected ratio ### Standardisation does not explain differences ### Life expectancy: hypothetical cohort #### Not one newborn's prediction ### Healthy life expectancy adjusts for nonfatal states ### DALYs: years of life lost plus years with disability ### QALYs mostly used in economic evaluation ### Weights do not measure a person's worth ## Attribution, surveillance, and data ### Attributable burden uses a counterfactual ### PAF: exposure prevalence, relative risk, causality ### Fractions can sum beyond one hundred per cent #### Risk factors overlap and interact ### Passive surveillance cheap but incomplete ### Active surveillance improves ascertainment ### Sentinel sites deep but may not represent ### Syndromic surveillance fast, less specific ### Testing changes create apparent trends ### Modelled certainty is not observed data ### Registration investment improves health ## Implementation science ### Evidence-based interventions into routine practice ### Effectiveness versus implementation ### Acceptability, adoption, fidelity, sustainability ### No access means no population benefit ### Fidelity preserves core causal functions #### Distinguish adaptation from active component loss ### Logic models make assumptions testable ## Programme evaluation and scale-up ### Process, outcome, impact ### Cluster and stepped randomised designs #### Rollout logistics do not ensure no bias ### Interrupted time series: level and trend ### Difference-in-differences: parallel trends ### Regression discontinuity uses a threshold ### Low reach can widen inequity ### Reach times effectiveness ### Scale-up is not a multiplied pilot ## Global health and assistance ### Transnational determinants, institutions, inequities ### Colonial history, trade, debt shape priorities ### Short external projects draw staff away ### Equitable partnership shares agenda and data ### Aid may be volatile and donor-directed ### Vertical, horizontal, diagonal approaches ### Parallel systems weaken sovereignty ## Emergencies and health security ### Violence or disaster with displacement ### Minimum service packages ### Accountability to affected people ### Vulnerability does not erase consent ### Prevent, detect, respond to cross-border threats ### Narrow wealthy-country focus undermines cooperation ### Transparency and fair countermeasures ## Planetary health ### Climate affects heat, food, water, air, infection ### Risks vary by place, work, age, housing ### Probability shifts across many events ### Mitigation versus adaptation ### Health system emissions ### Decarbonisation can align with health #### Removal without alternatives transfers risk ### Life-cycle assessment beyond visible waste ### One Health: humans, animals, environment #### Not every spillover is predictable ### Pollution burdens distributed unevenly ### Precaution is not evidence-free prohibition ### Polluters must not externalise costs ## Accountability ### Metrics reveal need and guide action ### Judge by reach, quality, equity, sustainability ### Responsibility across borders and generations