The STDRATE Procedure

This example computes the excess event risk fraction that is attributable to a specific chemical exposure for workers in a factory.

Suppose that the Factory data set contains the stratum-specific event information for exposure to a specific chemical agent. The variable Age is the grouping variable that forms the strata. The variables Event_E and Count_E indicate the number of events and number of workers for workers with the specific chemical exposure, respectively. The variables Event_NE and Count_NE indicate the number of events and number of workers for workers without the specific chemical exposure, respectively.

data Factory;
   input Age $ Event_E Count_E Event_NE Count_NE;
   datalines;
20-29   31  352  143  2626
30-39   57  486  392  4124
40-49   62  538  459  4662
50-59   50  455  337  3622
60-69   38  322  199  2155
70+      9   68   35   414
;

The following statements invoke the STDRATE procedure and compute the attributable risk and population attributable risk for the chemical exposure:

ods graphics on;
proc stdrate data=Factory
             refdata=Factory
             method=indirect(af)
             stat=risk
             plots(stratum=horizontal)
             ;
   population event=Event_E  total=Count_E;
   reference  event=Event_NE total=Count_NE;
   strata Age / stats;
run;
ods graphics off;

The “Standardization Information” table in Output 88.3.1 displays the standardization information.

The STATS option in the STRATA statement requests that the “Indirectly Standardized Strata Statistics” table in Output 88.3.2 display the strata information and the expected number of events at each stratum. The Expected Events column shows the expected numbers of events when the stratum-specific risks in the reference data set are applied to the corresponding numbers of workers in the study data set.

With ODS Graphics enabled and the specified STAT=RISK option, the default PLOTS=RISK option displays the stratum-specific risk estimates in the study and reference populations, as shown in Output 88.3.3. The STRATUM=HORIZONTAL global option in the PLOTS option displays the strata information on the horizontal axis. The plot displays the stratum-specific risk estimates in the “Indirect Standardized Strata Statistics” table in Output 88.3.2. In addition, confidence limits for the risk estimates in the study population and the overall crude risks for the two populations are also displayed

The METHOD=INDIRECT option requests that the “Standardized Morbidity/Mortality Ratio” table in Output 88.3.4 display the SMR, its confidence limits, and the test for the null hypothesis .

The “Standardized Morbidity/Mortality Ratio” table shows that SMR=1.259, the confidence limits do not contain the null value SMR=1, and the null hypothesis of SMR=1 is rejected at level from the normal test.

The “Indirectly Standardized Risk Estimates” table in Output 88.3.5 displays the standardized risks and related statistics.

The AF suboption in the METHOD=INDIRECT option requests that the “Attributable Fraction Estimates” table display the attributable risk and population attributable risk, as shown in Output 88.3.6

The attributable risk fraction 0.206 indicates that of all events in the chemical exposure group are attributed to the chemical exposure, and the population attributable risk fraction 0.028 indicates that about of all events in the total population are attributed to the chemical exposure.

The Attributable fraction can also be computed by using Mantel-Haenszel method.

Suppose that the Factory1 data set contains the stratum-specific event information for exposure to a specific chemical agent. The variable Age is the grouping variable that forms the strata, and the variable Exposure identifies workers with chemical exposure. The variables Event and Count indicate the number of events and number of workers, respectively.

data Factory1;
   input Exposure $ Age $ Event Count;
   datalines;
Yes  20-29   31   352
Yes  30-39   57   486
Yes  40-49   62   538
Yes  50-59   50   455
Yes  60-69   38   322
Yes  70+      9    68
No   20-29  143  2626
No   30-39  392  4124
No   40-49  459  4662
No   50-59  337  3622
No   60-69  199  2155
No   70+     35   414
;

The following statements invoke the STDRATE procedure and compute the attributable risk and population attributable risk for the chemical exposure:

proc stdrate data=Factory1
             method=mh(af)
             stat=risk
             effect
             ;
   population group(order=data)=Exposure event=Event total=Count;
   strata Age;
run;

The “Standardization Information” table in Output 88.3.7 displays the standardization information.

The “Mantel-Haenszel Standardized Risk Estimates” table in Output 88.3.8 displays the Mantel-Haenszel standardized risks and related statistics.

The EFFECT option requests that the “Risk Effect Estimates” table display the risk ratio statistic for the two directly standardized risks, as shown in Output 88.3.9.

The AF suboption in the METHOD=MH option requests that the “Attributable Fraction Estimates” table display the attributable risk and population attributable risk, as shown in Output 88.3.10

Similar to the results of using the SMR estimates, the attributable risk fraction (0.205) indicates that of all events in the chemical exposure group are attributed to the chemical exposure, and the population attributable risk fraction (0.028) indicates that about of all events in the total population are attributed to the chemical exposure.