In 1973, in cases where proxy responses were accepted, interviewers were required to "enter the column number of the person who did respond and mark the box that describes the reason why a proxy respondent was accepted" (i.e., "mentally incompetent," "physically incompetent," or "away from home during interview period").

The 1982 Manual placed particular emphasis on direct interviews for the preventive care questions. The 1982 Manual instructed interviewers to accept a proxy respondent for these questions only if

• The person is physically or mentally incapable of answering the questions OR

• The person is temporarily absent and will not return during the interview period OR

• The household has no telephone and you will not be in the same general area during the interview period.

In other cases,

• Do not accept a proxy-respondent if the person is not at home at the tome of your initial visit. Make at least three additional calls to make contact when the person is most likely to be home, as designated by the respondent. If the person cannot be reached on your third callback, accept a proxy respondent.

Other variables in the 1973 and 1982 surveys that report time since specific types of preventive care include EKGYRS (for electrocardiogram), GLAUCTESTYRS (for glaucoma test), and EYEXAMYRS (for vision test).

For 1973, responses to the question on years since last chest x-ray were topcoded at 10+ years in the NHIS public use files created by the National Center for Health Statistics. Single year responses up to 50 years for years since last chest x-ray are available for 1982, in CXRAYREX.

The 1973 Manual provided the following instruction to interviewers on how to record answers to the question for CXRAYRS:

By contrast, the 1982 Manual encouraged interviewers to press for exact responses and record fractions of years, as follows: