Ecologic studies JF Boivin Version 27 November 2006
Ecologic studies JF Boivin Version 27 November 2006 S: BOIVIN695Winter 2007Ecologic studies. ppt
Outline 1. Examples 2. Definition 3. Ecologic fallacy • Definition • Valid ecologic study • Rate difference varies • Reference rate varies 4. Ecologic confounder 5. Types of ecologic exposures 6. Rationale for ecologic studies 2
Percentage of children receiving measles-mumpsrubella immunization in second year of life and caseload of children with autism, by year of birth, California (Dales et al. , JAMA 2001) 3
4
(Goodman DC, et al. NEJM 2002) 5
(Goodman DC, et al. NEJM 2002) 6
Outline 1. Examples 2. Definition 3. Ecologic fallacy • Definition • Valid ecologic study • Rate difference varies • Reference rate varies 4. Ecologic confounder 5. Types of ecologic exposures 6. Rationale for ecologic studies 7
Ecologic study A study in which the units of analysis are populations or groups of people, rather than individuals. (Last. 2001) 8
Structure of an ecologic study: Counts Group 1 E+ E- D+ ? ? M 1+ D- ? ? M 1 - N 1+ N 1 - E+ E- D+ ? ? M 2+ D- ? ? M 2 - N 2+ N 2 - Group 2 9
Person-years Group 1 E+ E- D+ ? ? M 1+ PY PY 1+ PY 1 - PY 1 T E+ E- D+ ? ? M 2+ PY PY 2+ PY 2 - PY 2 T Group 2 10
Durkheim’s study Group 1 (provinces with protestant minority) Protestant Suicide PY ? 300, 000 Other ? 700, 000 10 1, 000 Group 2 (provinces with protestant majority) Protestant Suicide PY ? 800, 000 Other ? 200, 000 20 1, 000 11
Outline 1. Examples 2. Definition 3. Ecologic fallacy • Definition • Valid ecologic study • Rate difference varies • Reference rate varies across 4. Ecologic confounder 5. Types of ecologic exposures 6. Rationale for ecologic studies 12
Ecologic fallacy “…the mistaken assumption that a statistical association observed between two ecologic (group-level) variables… is equal to the association between the corresponding variables at the individual level…” (Encyclopedia of epidemiologic methods. 2000) 13
Ecologic fallacy “…arises when the disease rate in the unexposed (reference) population is correlated with exposure prevalence across groups or when the difference in rates between exposed and unexposed populations (biologic effect) varies across groups. ” (Encyclopedia of epidemiologic methods. 2000) 14
No ecologic bias Group 1 (Québec) E+ E- D+ 16 12 PY 8, 000 12, 000 28 20, 000 IE = 200/100, 000 Io = 100/100, 000 RD = 100/100, 000 RR = 2 % exposure = 8, 000/20, 000 = 40% Group rate = 28/20, 000 = 140/100, 000 Group 2 (Ontario) D+ PY E+ E- 24 8 12, 000 8, 000 32 20, 000 IE = 200/100, 000 Io = 100/100, 000 RD = 100/100, 000 RR = 2 % exposure = 12, 000/20, 000 = 60% Group rate = 32/20, 000 = 160/100, 000 Adapted from Rothman-Greenland Table 23 -2 15
No ecologic bias 200 190 RATE (per 100, 000) 180 170 Ontario 160 150 Québec 140 130 120 110 10 20 30 40 50 60 70 80 90 100 % EXPOSURE IRR = IE Io = 200/100, 000 = 2 100/100, 000 16
Ecologic bias (rate difference varies across groups) Group 1 (Québec) E+ E- D+ 20 13 PY 7, 000 13, 000 33 20, 000 IE = 286/100, 000 Io = 100/100, 000 RD = 186/100, 000 RR = 2. 86 % exposure = 7, 000/20, 000 = 35% Group rate = 33/20, 000 = 165/100, 000 Group 2 (Ontario) D+ PY E+ E- 20 7 13, 000 7, 000 27 20, 000 IE = 154/100, 000 Io 100/100, 000 = RD = 54/100, 000 RR = 1. 54 % exposure = 13, 000/20, 000 = 65% Group rate = 27/20, 000 = 135/100, 000 17
Ecologic bias 200 190 180 RATE (per 100, 000) 170 160 150 140 130 120 110 10 20 30 40 50 60 70 80 90 100 % EXPOSURE IRR = IE Io = 100/100, 000 = 0. 5 200/100, 000 18
Ecologic bias (reference rate varies across groups) Group 1 (Québec) E+ E- D+ 16 12 PY 8, 000 12, 000 28 20, 000 IE = 200/100, 000 Io = 100/100, 000 RD = 100/100, 000 RR = 2 % exposure = 8, 000/20, 000 = 40% Group rate = 28/20, 000 = 140/100, 000 Group 2 (Ontario) D+ PY E+ E- 40 6 16, 000 46 20, 000 IE = 250/100, 000 Io = 150/100, 000 RD = 100/100, 000 RR = 1. 67 % exposure = 16, 000/20, 000 = 80% Group rate = 46/20, 000 = 230/100, 000 19
Ecologic bias 250 RATE (per 100, 000) 200 150 100 0 10 20 30 40 50 60 70 80 90 100 % EXPOSURE IRR = IE Io 275/100, 000 = = 5. 5 50/100, 000 20
(Koepsell & Weiss) 21
Outline 1. Examples 2. Definition 3. Ecologic fallacy • Definition • Valid ecologic study • Rate difference varies • Reference rate varies 4. Ecologic confounder 5. Types of ecologic exposures 6. Rationale for ecologic studies 22
No ecologic bias Country 1 Country 2 Country 3 Covariate Exposure PY Rate Yes 3000 600 4000 600 No 12000 500 8000 500 6000 500 RR 1. 2 No Yes 2000 600 4000 6000 600 No 8000 500 9000 500 RR 1. 2 Sum Yes 5000 600 8000 600 10000 600 No 20000 500 16000 500 15000 500 RR 1. 2 (rates per 100, 000 person-year) (Note: no individual-level confounding) Ecologic analysis % exposed 20% 33% 40% % covariate 60% 50% 40% Overall disease rate 520 533 540 Crude RR = 1. 2 Adjusted RR = 1. 2 Crude is valid! 23
Crude ecologic bias No stratum-specific ecologic bias Country 1 Country 2 Country 3 Covariate Exposure PY Rate Yes 3000 600 4000 600 No 12000 500 8000 500 6000 500 RR 1. 2 No Yes 2000 200 4000 200 6000 200 No 8000 100 9000 100 RR 2 2 2 Sum Yes 5000 440 8000 400 10000 360 No 20000 340 16000 300 15000 260 RR 1. 3 1. 4 (rates per 100, 000 person-year) (Note: no individual-level confounding) Ecologic analysis % exposed 20% 33% 40% % covariate 60% 50% 40% Overall disease rate 360 333 300 Crude RR = 0. 3 Adjusted RR = 1. 3 Adjusted is valid! 24
Ecologic bias (crude and stratum-specific) Country 1 Country 2 Country 3 Covariate Exposure PY Rate Yes 8000 500 13000 1500 14000 1000 No 12000 500 12000 1500 6000 1000 RR 1 1 1 No Yes 2000 100 2000 300 6000 200 No 28000 100 23000 300 24000 200 RR 1 1 1 Sum Yes 10000 420 15000 1340 20000 760 No 40000 220 35000 711 30000 360 RR 1. 9 2. 1 (rates per 100, 000 person-year) Ecologic analysis % exposed 20% 30% 40% % covariate 40% 50% 40% Overall disease rate 260 900 520 Crude RR = 8. 6 Adjusted RR = 8. 6 No valid estimate available! 25
Outline 1. Examples 2. Definition 3. Ecologic fallacy • Definition • Valid ecologic study • Rate difference varies • Reference rate varies 4. Ecologic confounder 5. Types of ecologic exposures 6. Rationale for ecologic studies 26
Aggregate exposure Attributes of individuals that are summarized at the group level • • Proportion of smokers Median family income Proportion of protestants Prevalence of subjects who are immune to measles Scientific interest may lie in: • Individual effect • Contextual effect 27
Intrinsically population-level exposure Attributes of groups for which no distinct analog exists at the individual level • • • Population density Law Health-care system Social disorganization Income discrepancy Everybody is exposed! 28
29
• Neighborhood social class as aggregate of individual social classes Can differ from study subjects’ social class • Neighborhood social class as contextual variable Same contextual variable for all subjects The variable is ecological, but the study is not! 30
31
Outline 1. Examples 2. Definition 3. Ecologic fallacy • Definition • Valid ecologic study • Rate difference varies across groups • Reference rate varies across groups 4. Ecologic confounder 5. Types of ecologic exposures 6. Rationale for ecologic studies 32
Rationale for ecologic studies 1. Low cost and convenience 2. Measurement limitation of individual-level studies 3. Design limitations of individuallevel studies • Koepsell and Weiss, Figure 12. 1 4. Simplicity of analysis and presentation ? Interest in ecologic effects 33
(Koepsell & Weiss) 34
- Slides: 34