3: Indirect standardization Tables 5.25 and 5.26 are data representing the number of inpatient complications due to myocardial infarction during 2003 in Jefferson County, the service area for East Bank Regional Hospital, and Washington County, the service area for Westside Memorial Hospital. Data are presented in age categories. a. The following data was presented in class in the PowerPoint slides and is given here. The question asks: Using these data, calculate age-specific inpatient MI complication rates for Jefferson and Washington counties, per 1,000 population, and fill in Tables 5.25 and 5.26. Table 5.25 Jefferson County The table asks that you calculate the Crude Rate. Since we are calculating a rate, the math process is to divide the Complications by the Population. You then multiply by 1000 to get the rate per 1000. Age Number of MI Complications Population Crude Rate (per 1,000) 40-49 196 80,300 2.44 50-59 169 97,400 1.73 60-69 623 52,500 11.87 70 and older 1,920 37,250 51.54 Totals 2,908 267,450 10.87 Table 5.26 Washington County Age Number of MI Complications Population Crude Rate (per 1,000) 40-49 82 34.500 2.38 50-59 53 27,600 1.92 60-69 436 7,600 57.37 70 and older 125 5,715 21.87 Totals 696 75,415 9.23 b. The following graph was provided in class and is repeated here. The question asks: Compare age-specific complication rates between Jefferson and Washington Counties. Which county has the higher age-specific complication rates? c. Discuss the general pattern of the change in complication rate with age. d. Calculate the TOTAL crude complication rate for Jefferson and Washington counties for 2003 (you already did this in the preceding tables). Crude complication rate per 1,000 in 2003 (see the tables above) Jefferson County = 10.87 Washington County = 9.23 e. What is the ratio of the TOTAL crude complication rate in Jefferson County to that in Washington County? f. Do you draw the same conclusion about the complication rates in Jefferson versus Washington County when you compare age-specific rates as you do when you compare TOTAL crude complication rates? Explain. g. Part g is not required for this exam. h. The data presented in Tables 5.27 and 5.28 indicate that the population distributions of Jefferson and Washington Counties are quite different. Calculate the percent of the population in each age group in both populations, and fill in the tables. Table 5.27 Jefferson County Age Population 40-49 80,300 50-59 97,400 60-69 52,500 70 and older 37,250 Totals 267,450 Percent Table 5.28 Washington County Age Population 40-49 34,500 50-59 27,600 60-69 7,600 70 and older 5,715 Totals 75,415 Percent i. Based on the preceding calculations, what is the difference between the two counties with respect to their age distributions (which county has the younger population)? Age Jefferson County Washington County < 59 > 59 j. Based on questions h and i, the differences in age distribution require adjustment of complication rates for age differences before single, summary rates between the two counties can be compared. Table 5.29: Standard Population Crude MI Complication Rates Age Group Standard Population MI Rates (U.S. 1990) 40-49 .0035 50-59 .0025 60-69 .0075 70 and older .0095 Given Table 5.29, perform indirect adjustment (indirect standardization) of crude complication rates. Direct standardization Estimate that people in the entire reference population under study are being impacted at the same rate as in your target population, or in Table 5.4, that the study population is visiting an outpatient department at the same rate as the entire population of the U.S. Indirect standardization Estimate that the number of people in your study population would be impacted at the same rate as the entire population. This allows for the calculation of a Standard Mortality or Morbidity Ratio (SMR), a better estimator of mortality (see page 108). The first step is to select a standard population and its MI rates. The standard population selected is the 2000 U.S. population (see Table 5.30). The second step in the indirect adjustment of rates is to calculate the expected number of complications in each age group. This is accomplished by multiplying the number of persons in each age category in the study populations (Jefferson and Washington Counties) by the rates of MI in each age category in the standard population (2000 U.S.). Fill in Tables 5.30 and 5.31. Calculate the standardized morbidity ratio for Jefferson and Washington Counties, by dividing the observed cases by the expected cases: Table 5.30 Jefferson County Age Standard Population MI Rates (U.S. 1990) Jefferson County Population 40-49 .0035 80,300 50-59 .0025 97,400 60-69 .0075 52,500 70 and older .0095 37,250 Totals 267,450 Expected Complications Table 5.31 Washington County Age Standard Population MI Rates (U.S. 1990) Washington County Population 40-49 .0035 34,500 50-59 .0025 27,600 60-69 .0075 7,600 70 and older .0095 5,715 Totals Expected Complications 75,415 SMR = observed cases expected cases A Standard Mortality Ratio (SMR) is calculated as the ratio of actual expected deaths or morbidity in two comparison populations. If the ratio is more than 1.0, it means that actual deaths exceeded expected deaths, or that the comparison population was dying at rates in excess of the standard population. A ratio less than 1.0 implies that the comparison group had fewer deaths than expected, or that the rate was less than that of the standard population group. If the ratio is equal to 1.0, then there is no difference. Jefferson County SMR = observed cases (Total Number of complications from Table 5.25) divided by expected cases (Total number of complications from Table 5.30) Jefferson County SMR = observed cases expected cases Washington County SMR = observed cases expected cases k. Discuss the difference in SMR for Jefferson and Washington Counties. What does this difference indicate to you?