Notes on Risk: Concepts, Assessment, and Management

Definition of Risk

  • Risk is the probability of suffering harm from a hazard.
    • Harm is defined as injury; at environmental or system levels, harms can be more complex but still framed as injury or negative outcomes.
    • Environmental safety systems (e.g., in cars) are designed to reduce risk of injury, not completely prevent it.
  • In everyday thinking, risk is linked to statistical and mathematical concepts, but not everyone is comfortable with math.
  • The lecture emphasizes connecting risk concepts to real-world environments and long-term implications (health, environment, economy).

How Risk is Measured and What Harm Means

  • Harm = injury; risk of harm from a hazard = probability of injury after exposure to a hazard.
  • In vehicles, airbags, cushioned dashboards, seat belts all reduce injury risk but do not eliminate it.
  • Environmental/system-level risk involves figuring out how likely some injury or damage will occur within a system or ecosystem.

Disease Risk, Transmission, and Reducing Risk

  • Measles example to illustrate transmission and risk:
    • Transmission is via aerosol (sneezing, coughing, etc.).
    • The transmission probability (R0, reproduction number) varies by disease.
  • Reproduction number (R0): average number of people that one infected person will infect in a fully susceptible population.
    • COVID-19: R0 ≈ 1.4R_0 \,\approx\, 1.4
    • Measles: R0 ≈ 12R_0 \,\approx\, 12
  • Implication: higher R0 means faster, more widespread transmission if unmitigated.
  • Mitigation measures to reduce risk of widespread transmission:
    • Vaccinations (creates immunity in population).
    • Masks (reduces transmission rates).
    • Isolation of exposed individuals.
    • Note: These measures reduce risk but do not provide absolute protection (not a "full body condom").
  • Conceptual takeaway: risk reduction often requires multiple interventions to lower transmission and harm.

Economic and Environmental Risks

  • Environmental degradation and climate change lead to economic losses and higher risk in daily life.
  • Example: beef prices around {"$25"} per pound due to droughts and reduced cattle production linked to climate factors.
  • Broader reality: heat, drought, and other climate impacts translate into increased costs and risk in food systems and livelihoods.
  • Sometimes framed as: risk of death or serious harm is ever-present, and adaptation or avoidance strategies can reduce that risk (contextualized in environmental terms).

Everyday Risks and How We Express Them

  • Risk is commonly expressed as a probability: the likelihood of an event occurring.
  • Geographic context matters for certain risks (e.g., shark attacks are more likely along the Gulf Coast than Missouri).
  • Quick examples of daily risk considerations:
    • Health risks like smoking and secondhand smoke contribute to long-term harm.
    • Daily accident risk from car travel can be significant even if not immediately obvious.
    • Planes often evoke fear due to dramatic outcomes, even though statistically less likely than driving.
  • The speaker notes that people tend to misjudge which risks are most dangerous due to salience, fear, or media coverage.

Health Risks: Smoking, Car Crashes, and Plane Crashes

  • Smoking and lifetime cancer risk (lung cancer) for a pack-a-day smoker:
    • Probability: P( ext{lung cancer} \mid \text{smoking}) \,=\, \frac{1}{250}</li><li>Note:secondhandsmokeexposureandforestfiresmokealsocontributetoriskbutarenotalwayscapturedinthemainstatistic.</li><li>Riskreduction:quittingsmokinggreatlylowersriskovertime.</li></ul></li><li>Carcrashes:<ul><li>Probabilityofdyinginacarcrash:</li> <li>Note: secondhand smoke exposure and forest fire smoke also contribute to risk but are not always captured in the main statistic.</li> <li>Risk reduction: quitting smoking greatly lowers risk over time.</li></ul></li> <li>Car crashes:<ul> <li>Probability of dying in a car crash:P(\text{death in a car crash}) \,\approx\, \frac{1}{93}</li></ul></li><li>Planecrashes:<ul><li>Probabilityofdyinginaplanecrash:</li></ul></li> <li>Plane crashes:<ul> <li>Probability of dying in a plane crash:P(\text{death in plane crash}) \,\approx\, \frac{1}{11{,}000{,}000}</li><li>Despitethedramaticnatureofairdisasters,thestatisticalriskpertripisverysmallcomparedtocartravel,butperceptioncanbeheightenedbymediacoverage.</li></ul></li></ul><h3id="floodsclimateextremesandthethousandyearfloodconcept">Floods,ClimateExtremes,andtheThousand−YearFloodConcept</h3><ul><li>Thousand−yearflood:acommonlyusedtermforafloodeventwithareturnperiodofabout1000years.<ul><li>Definition:aneventlikelytooccuronceinathousandyearsonaverage.</li><li>Caveat:climatevariabilitymeanssucheventscanoccurinconsecutiveyears;statisticsdon’tperfectlypredictexacttiming.</li><li>Example:upstreamMissouriRiverstudyshowedmultiple1000−yearfloodsina40−yearspan(notasingle−yearexpectation).</li></ul></li><li>Excessiveheat:<ul><li>From1979to2022,deathratesduetoheathaverisen;insomeassessments,ratesrosefromroughly</li> <li>Despite the dramatic nature of air disasters, the statistical risk per trip is very small compared to car travel, but perception can be heightened by media coverage.</li></ul></li> </ul> <h3 id="floodsclimateextremesandthethousandyearfloodconcept">Floods, Climate Extremes, and the Thousand-Year Flood Concept</h3> <ul> <li>Thousand-year flood: a commonly used term for a flood event with a return period of about 1000 years.<ul> <li>Definition: an event likely to occur once in a thousand years on average.</li> <li>Caveat: climate variability means such events can occur in consecutive years; statistics don’t perfectly predict exact timing.</li> <li>Example: upstream Missouri River study showed multiple 1000-year floods in a 40-year span (not a single-year expectation).</li></ul></li> <li>Excessive heat:<ul> <li>From 1979 to 2022, death rates due to heat have risen; in some assessments, rates rose from roughly0.5togreaterthanto greater than5(unitsnotspecifiedintranscript).</li><li>2021and2022wereamongthehottestyearsonrecord;heatdomesandclimatechangeamplifyhealthandecologicalrisks.</li><li>Olderpopulationsareathigherriskfromexcessiveheatduetoreducedthermoregulation,healthcomorbidities,andotherfactors.</li></ul></li><li>Real−worldimplication:climatechangecontributestohigherpopulation−levelriskthroughheatstress,drought,andrelatedhealthandeconomicimpacts.</li></ul><h3id="possibilityvsprobability">PossibilityvsProbability</h3><ul><li>Possibility:anon−quantifiedpotentialforsomethingtooccur(conceptual).<ul><li>Example:smokersmaydeveloplungcancerisapossibility,butnotaquantifiedrisk.</li></ul></li><li>Probability:anumericalmeasureoflikelihood;providesastatisticalwaytoevaluaterisk.<ul><li>Notation:probabilityisdenotedbyP(⋅).</li><li>Asascientist,assigningnumbers(probabilities)toriskfactorssupportsprioritizationanddecision−making.</li></ul></li></ul><h3id="riskassessmentraestimatingrisk">RiskAssessment(RA):EstimatingRisk</h3><ul><li>Definition:anexplicit,scientificprocessofstatisticallyestimatingrisk.</li><li>Purpose:tocompareandcontrastdifferentriskfactorsandtotriagewhichriskstoaddresswithlimitedresources.</li><li>Coreidea:notallriskscanbemitigated;focusonthosewiththegreatestimpactonhealth,safety,orenvironment.</li><li>Exampleframework(CO2andtemperature):<ul><li>IncreasedCO2levelsleadtoincreasedatmospherictemperatures(assessedrisk).</li><li>Quantifiedscenarioprobabilitiesgivenpolicychoices:</li><li>Ifcurrentpoliciescontinue:(units not specified in transcript).</li> <li>2021 and 2022 were among the hottest years on record; heat domes and climate change amplify health and ecological risks.</li> <li>Older populations are at higher risk from excessive heat due to reduced thermoregulation, health comorbidities, and other factors.</li></ul></li> <li>Real-world implication: climate change contributes to higher population-level risk through heat stress, drought, and related health and economic impacts.</li> </ul> <h3 id="possibilityvsprobability">Possibility vs Probability</h3> <ul> <li>Possibility: a non-quantified potential for something to occur (conceptual).<ul> <li>Example: smokers may develop lung cancer is a possibility, but not a quantified risk.</li></ul></li> <li>Probability: a numerical measure of likelihood; provides a statistical way to evaluate risk.<ul> <li>Notation: probability is denoted by P(⋅).</li> <li>As a scientist, assigning numbers (probabilities) to risk factors supports prioritization and decision-making.</li></ul></li> </ul> <h3 id="riskassessmentraestimatingrisk">Risk Assessment (RA): Estimating Risk</h3> <ul> <li>Definition: an explicit, scientific process of statistically estimating risk.</li> <li>Purpose: to compare and contrast different risk factors and to triage which risks to address with limited resources.</li> <li>Core idea: not all risks can be mitigated; focus on those with the greatest impact on health, safety, or environment.</li> <li>Example framework (CO2 and temperature):<ul> <li>Increased CO2 levels lead to increased atmospheric temperatures (assessed risk).</li> <li>Quantified scenario probabilities given policy choices:</li> <li>If current policies continue:P(\Delta T \ge 2.8^\u00b0C) = 0.66(i.e.,66<li>Ifnewpoliciesareimplemented:(i.e., 66% chance of at least 2.8°C warming).</li> <li>If new policies are implemented:P(\Delta T \ge 2.5^\u00b0C) = 0.66(66<li>Conversionreminder:temperatureriseinCelsiuscanbeconvertedtoFahrenheitby(66% chance of at least 2.5°C warming).</li> <li>Conversion reminder: temperature rise in Celsius can be converted to Fahrenheit by\Delta T{\mathrm{F}} = \Delta T{\mathrm{C}} \times \tfrac{9}{5},so2.8°C≈5.04°Fand2.5°C≈4.50°F.</li></ul></li><li>Practicaltakeaway:RAprovidesastructuredbasistodecidewhichriskstoaddressandhowintensivelytointervene.</li><li>Noteondataquality:policy−adjustedprobabilitiesareestimates,andsmalldifferences(e.g.,2.5°Cvs2.8°C)canstillbemeaningfulforecosystemsandhumanhealth.</li></ul><h3id="riskmanagementrmactingonrafindings">RiskManagement(RM):ActingonRAFindings</h3><ul><li>Definition:RMusestheoutcomesofRAtodecidehowtomitigateorreduceariskfactor.</li><li>Limitation:Youcannoteliminateallrisk;someresidualriskremainsevenaftermitigation.</li><li>Examplesofmitigationstrategiesincludevaccination,masking,isolation,engineeringcontrols,policies,andbehaviorchange.</li><li>Philosophical/biologicalreminder:riskisintrinsictolife;someriskisunavoidablefromthemomentofbirth(oxygenisessentialbutcanbetoxicatcellularlevelsifmismanaged).</li><li>Practicalimplication:implementlayereddefenses(defenseindepth)toreduceriskacrossmultiplepathways.</li></ul><h3id="connectionsimplicationsandtakeaways">Connections,Implications,andTakeaways</h3><ul><li>Riskblendseverydayintuitionwithquantitativemethods(probability,statistics,andmodeling).</li><li>UnderstandingR0helpsexplainwhysomediseasesspreadmoreeasilyandwhycertaininterventionsareprioritized.</li><li>Climateandenvironmentalrisksrequireprobabilisticreasoningandriskmanagementtoreducehealthandeconomicimpacts.</li><li>Publichealthdecisions(vaccinationprograms,masks,isolation)aregroundedinriskassessmentandriskmanagementframeworks.</li><li>Ethicalandpracticalconsiderations:balancingindividualfreedomswithpopulation−levelprotection,communicatingriskwithoutsensationalism,andaddressingdisparitiesinvulnerability(age,health,socioeconomicstatus).</li></ul><h3id="quickreferenceformulasandkeynumbers">QuickReferenceFormulasandKeyNumbers</h3><ul><li>Risk(harm)definition:<ul><li>, so 2.8°C ≈ 5.04°F and 2.5°C ≈ 4.50°F.</li></ul></li> <li>Practical takeaway: RA provides a structured basis to decide which risks to address and how intensively to intervene.</li> <li>Note on data quality: policy-adjusted probabilities are estimates, and small differences (e.g., 2.5°C vs 2.8°C) can still be meaningful for ecosystems and human health.</li> </ul> <h3 id="riskmanagementrmactingonrafindings">Risk Management (RM): Acting on RA Findings</h3> <ul> <li>Definition: RM uses the outcomes of RA to decide how to mitigate or reduce a risk factor.</li> <li>Limitation: You cannot eliminate all risk; some residual risk remains even after mitigation.</li> <li>Examples of mitigation strategies include vaccination, masking, isolation, engineering controls, policies, and behavior change.</li> <li>Philosophical/biological reminder: risk is intrinsic to life; some risk is unavoidable from the moment of birth (oxygen is essential but can be toxic at cellular levels if mismanaged).</li> <li>Practical implication: implement layered defenses (defense in depth) to reduce risk across multiple pathways.</li> </ul> <h3 id="connectionsimplicationsandtakeaways">Connections, Implications, and Takeaways</h3> <ul> <li>Risk blends everyday intuition with quantitative methods (probability, statistics, and modeling).</li> <li>Understanding R0 helps explain why some diseases spread more easily and why certain interventions are prioritized.</li> <li>Climate and environmental risks require probabilistic reasoning and risk management to reduce health and economic impacts.</li> <li>Public health decisions (vaccination programs, masks, isolation) are grounded in risk assessment and risk management frameworks.</li> <li>Ethical and practical considerations: balancing individual freedoms with population-level protection, communicating risk without sensationalism, and addressing disparities in vulnerability (age, health, socioeconomic status).</li> </ul> <h3 id="quickreferenceformulasandkeynumbers">Quick Reference Formulas and Key Numbers</h3> <ul> <li>Risk (harm) definition:<ul> <li>\text{Risk} = P(\text{harm} \mid \text{hazard})</li></ul></li><li>Reproductionnumbers:<ul><li>COVID−19:</li></ul></li> <li>Reproduction numbers:<ul> <li>COVID-19:R_0 \approx 1.4</li><li>Measles:</li> <li>Measles:R_0 \approx 12</li></ul></li><li>Diseasetransmissionprobabilities(examples):<ul><li>Lungcancerriskforapack−a−daysmoker:</li></ul></li> <li>Disease transmission probabilities (examples):<ul> <li>Lung cancer risk for a pack-a-day smoker:P(\text{lung cancer} \mid \text{smoking}) = \frac{1}{250}</li><li>Carcrashfatalityrisk:</li> <li>Car crash fatality risk:P(\text{death in auto crash}) \approx \frac{1}{93}</li><li>Planecrashfatalityrisk:</li> <li>Plane crash fatality risk:P(\text{death in plane crash}) \approx \frac{1}{11{,}000{,}000}</li></ul></li><li>Floodingandreturnperiod:returnperiod</li></ul></li> <li>Flooding and return period: return periodT = 1000\text{ years}→annualprobability→ annual probabilityP(\text{flood in a given year}) \approx \frac{1}{1000} = 0.001</li><li>Temperaturechangeprobabilityscenarios(RA):<ul><li></li> <li>Temperature change probability scenarios (RA):<ul> <li>P(\Delta T \ge 2.8^\u00b0C \mid \text{current policies}) = 0.66</li><li></li> <li>P(\Delta T \ge 2.5^ 0C \mid \text{new policies}) = 0.66</li></ul></li><li>Temperatureconversionreminder:<ul><li></li></ul></li> <li>Temperature conversion reminder:<ul> <li>\Delta T{\mathrm{F}} = \Delta T{\mathrm{C}} \times \tfrac{9}{5}</li><li>Example:</li> <li>Example:2.8^{\circ}\mathrm{C} \approx 5.04^{\circ}\mathrm{F},,2.5^{\circ}\mathrm{C} \approx 4.50^{\circ}\mathrm{F}</li></ul></li><li>Populationallergytovaccineingredients:approximately</li></ul></li> <li>Population allergy to vaccine ingredients: approximatelyP(\text{allergic to vaccine ingredients}) \approx 0.01$$ (1%)

    Endnote

    • Questions are welcome; risk concepts become clearer when you apply them to concrete scenarios and data.