Codes and Frequencies
For sample adults and sample children, DVINT reports the approximate amount of time since the individual last saw or talked to a medical doctor (1969-1981), a medical doctor or assistant (1982-1996) or a doctor or other health care professional (1997-forward). In 1963, 1964, and 1967, DVINT was asked of persons in two quarters and in 1968-1996, it was asked of all persons.
Prior to 1982, the information for DVINT was collected through a series of questions, first dealing with doctor visits during the past two weeks and then, for those who had not seen or talked to a doctor during the preceding two weeks, through a general question about time since seeing or talking to a doctor (or doctor's assistant, for 1982 forward). Beginning in 1982, the information for DVINT was collected only through a general question about time since seeing or talking to a doctor or doctor's assistant (1982-1996) or to a doctor or other health care professional (1997-forward).
There were important changes in both the question wording and the instructions to interviewers. Particularly significant are differences in the meaning of DVINT for 1969-1981 versus 1982-1996.
Changes between 1969 and 1981 were relatively minor.
The original question wording (for 1969-1974) was, "About how long has it been since [person] saw or talked to a medical doctor?" For 1975-81, this wording was retained, but the line, "(Include doctors seen while a patient in a hospital)" was added.
The Field Representative's Manual for 1969-73 specified, "The term 'doctor' covers only medical doctors (MD) and osteopathic physicians (OD)." Moreover, the Manual directed interviewers, "Do not count dentists, optometrists or other practitioners (such as chiropractors) who do not have an MD degree as medical doctors." For 1974-1981, the definition of medical doctor remained the same, but the Field Representative's Manual for those years instructed the interview not to "make a special inquiry about the kind of doctor consulted or tell the respondent the survey definition of who is considered a doctor." Instead, if the respondent volunteered information about having seen some other sort of health professional (e.g., a chiropractor), interviewers were directed to repeat the question, emphasizing "medical" (e.g., "Did you see or talk to a medical doctor?")
Beginning in 1982, the scope of the question broadened considerably.
Specifically, from 1982 to 1996, interviewers asked, "About how long has it been since [person] last saw or talked to a medical doctor or assistant? Include doctors seen while a patient in a hospital." Beginning in 1997, the question was, "About how long has it been since [you/anyone in the family] last saw or talked to a doctor or other health care professional about [your own/sample child's] health? Include doctors seen while [he/she was] a patient in a hospital." From 1982 forward, then, contact with a health care professional such as a nurse practitioner in a clinic or doctor's office counted, which was not the case earlier. Researchers should therefore be cautious about comparing results from before and after 1982.
A change in the directions to interviewers in the Field Representative's Manual for 1982 forward also made DVINT more inclusive in later years.
Beginning in 1982, the Manual stated that the terms "medical doctor" and "assistant" should be "respondent defined." The one exception was that "visits to dentists or oral surgeons" were not to be included. Thus, for example, a reported visit to a chiropractor would have been excluded for 1969-81 and included for 1982 forward.
The intervals used to record respondents' answers changed over time. To maximize comparability, IPUMS NHIS uses a composite coding system for DVINT.
The first digit identifies time periods which can be identified in all years: Never (code 1), Less than 1 year (code 2), 1 year or more (code 3), 5 years or more (code 4), and unknown (code 9). The second and third digits identify periods specified for only a subgroup of years. Thus, for example, the longest time period coded varied across years: more than 3 years (for 1997-1998, code 310); 5 or more years (for 1970-1996 and 1999-2018, code 400); and more than 10 years (1969 and 2019 forward, code 402).
The NHIS questionnaire was substantially redesigned in 2019 to introduce a different data collection structure and new content. For more information on changes in terminology, universes, and data collection methods beginning in 2019, please see the user note.
- 1963 1964 1967: Persons in two quarters.
- 1968-1996: All persons.
- 1997-2018: Sample adults age 18+ and sample children ages 0-17.
- 2019-2020: Sample adults age 18+ and sample children ages 0-17.
- 1963-1964, 1967-2020