SmokingNow not correctly coded (or documented)

[nicholasjhorton]

tally(SmokeNow ~ Smoke100, data=NHANES) Smoke100 SmokeNow No Yes No 0 1745 0 Yes 0 1466 0

4024 0 2765

Note that if Smoke100=="No" then "SmokeNow" should be "No". Either this should be fixed (or the documentation: which says that SmokeNow is:

Study participant currently smokes cigarettes regularly. Reported for participants aged 20 years or older as Yes or No.

(i.e., it doesn't say that they are lifetime smokers). I'd much prefer the former fix (recode the missing SmokeNow as "No") than the latter.

[rpruim]

It looks to me like what we have is correct (see tables below) in that it matches the tables on the NHANES site. The trick is this, if a subject answers No to have you smoked 100 cigarettes, they are not asked whether they are currently smokers. Likely most of them are not current smokers, but it is possible that someone just started, I guess.

> tally( ~Smoke100 + SmokeNow | SurveyYr, data = NHANESraw, margins=TRUE)
, , SurveyYr = 2009_10

        SmokeNow
Smoke100    No   Yes  <NA> Total
   No        0     0  3352  3352
   Yes    1520  1346     0  2866
   <NA>      0     0  4319  4319
   Total  1520  1346  7671 10537

, , SurveyYr = 2011_12

        SmokeNow
Smoke100    No   Yes  <NA> Total
   No        0     0  3184  3184
   Yes    1259  1108     2  2369
   <NA>      0     0  4203  4203
   Total  1259  1108  7389  9756

The HANES site has this for 2011-12:

[rpruim]

I've added the following example:

> if (require(mosaic)) {
+   nhanes <- 
+     NHANES %>% 
+     mutate(
+       SmokingStatus = derivedFactor(
+         Current = SmokeNow == "Yes",
+         Former = SmokeNow == "No",
+         Never  = Smoke100 == "No"
+         )
+       )
+    tally( ~SmokingStatus, data = nhanes )
+ }

Current  Former   Never    <NA> 
   1466    1745    4024    2765 

and updated documentation to make the question flow clear.