Outcome-based doctor-patient interaction analysis

I. Comparison of techniques

Thomas Inui, William B. Carter, Walter A. Kukull, Virginia H. Haigh

Research output: Contribution to journalArticle

120 Citations (Scopus)

Abstract

Interactional analysis (IA) systems have been devised and applied to doctor- patient dialogues to describe encounters and to relate process to outcomes. Prior work in this area has been typified by the use of a single taxonomy for classifying verbal behaviors and limited outcomes (compliance and/or satisfaction). We applied three different IA systems (Bales, Roter’s modified Bales with affective ratings, and Stiles’ “Verbal Response Modes”) to 101 new-patient visits to a general medical clinic for which multiple outcomes had been determined: Several measures of patient knowledge of problems at conclusion of visit; patient compliance with drugs (over the ensuing three months); and patient satisfaction with the visit (perceived technical, interpersonal and communication quality). Within IA systems, cross tabulations and multiple regressions were performed to relate encounter events to outcomes. Across IA systems, multiple regression R2and R2adjusted (R2α) for the number of independent variables entering were used to characterize strength of relationships. Roter’s IA system showed stronger relationships to outcomes of knowledge (41% R2, 27% R2α) and compliance (44% R2, 28% R2α) than did Bales’ or Stiles’ systems. R2for patient satisfaction was identical for Bales and Roter (35%), and greater than R2 for Stiles (14%). We conclude that choice of IA system for research or teaching purposes should be based on behaviors and outcomes of particular interest and importance to the user. Based on audio-review of tapes, Roter’s approach is less time-consuming and may perform as well as more complex systems requiring transcript analysis.

Original languageEnglish (US)
Pages (from-to)535-549
Number of pages15
JournalMedical Care
Volume20
Issue number6
StatePublished - 1982
Externally publishedYes

Fingerprint

systems analysis
interaction
Patient Satisfaction
Compliance
Verbal Behavior
Patient Compliance
regression
language behavior
Teaching
Communication
taxonomy
bismuth subnitrate
rating
dialogue
Research
Pharmaceutical Preparations
drug
event
communication

ASJC Scopus subject areas

  • Public Health, Environmental and Occupational Health
  • Nursing(all)
  • Health(social science)
  • Health Professions(all)

Cite this

Inui, T., Carter, W. B., Kukull, W. A., & Haigh, V. H. (1982). Outcome-based doctor-patient interaction analysis: I. Comparison of techniques. Medical Care, 20(6), 535-549.

Outcome-based doctor-patient interaction analysis : I. Comparison of techniques. / Inui, Thomas; Carter, William B.; Kukull, Walter A.; Haigh, Virginia H.

In: Medical Care, Vol. 20, No. 6, 1982, p. 535-549.

Research output: Contribution to journalArticle

Inui, T, Carter, WB, Kukull, WA & Haigh, VH 1982, 'Outcome-based doctor-patient interaction analysis: I. Comparison of techniques', Medical Care, vol. 20, no. 6, pp. 535-549.
Inui, Thomas ; Carter, William B. ; Kukull, Walter A. ; Haigh, Virginia H. / Outcome-based doctor-patient interaction analysis : I. Comparison of techniques. In: Medical Care. 1982 ; Vol. 20, No. 6. pp. 535-549.
@article{412b8dd1f018409bbb4034d802934057,
title = "Outcome-based doctor-patient interaction analysis: I. Comparison of techniques",
abstract = "Interactional analysis (IA) systems have been devised and applied to doctor- patient dialogues to describe encounters and to relate process to outcomes. Prior work in this area has been typified by the use of a single taxonomy for classifying verbal behaviors and limited outcomes (compliance and/or satisfaction). We applied three different IA systems (Bales, Roter’s modified Bales with affective ratings, and Stiles’ “Verbal Response Modes”) to 101 new-patient visits to a general medical clinic for which multiple outcomes had been determined: Several measures of patient knowledge of problems at conclusion of visit; patient compliance with drugs (over the ensuing three months); and patient satisfaction with the visit (perceived technical, interpersonal and communication quality). Within IA systems, cross tabulations and multiple regressions were performed to relate encounter events to outcomes. Across IA systems, multiple regression R2and R2adjusted (R2α) for the number of independent variables entering were used to characterize strength of relationships. Roter’s IA system showed stronger relationships to outcomes of knowledge (41{\%} R2, 27{\%} R2α) and compliance (44{\%} R2, 28{\%} R2α) than did Bales’ or Stiles’ systems. R2for patient satisfaction was identical for Bales and Roter (35{\%}), and greater than R2 for Stiles (14{\%}). We conclude that choice of IA system for research or teaching purposes should be based on behaviors and outcomes of particular interest and importance to the user. Based on audio-review of tapes, Roter’s approach is less time-consuming and may perform as well as more complex systems requiring transcript analysis.",
author = "Thomas Inui and Carter, {William B.} and Kukull, {Walter A.} and Haigh, {Virginia H.}",
year = "1982",
language = "English (US)",
volume = "20",
pages = "535--549",
journal = "Medical Care",
issn = "0025-7079",
publisher = "Lippincott Williams and Wilkins",
number = "6",

}

TY - JOUR

T1 - Outcome-based doctor-patient interaction analysis

T2 - I. Comparison of techniques

AU - Inui, Thomas

AU - Carter, William B.

AU - Kukull, Walter A.

AU - Haigh, Virginia H.

PY - 1982

Y1 - 1982

N2 - Interactional analysis (IA) systems have been devised and applied to doctor- patient dialogues to describe encounters and to relate process to outcomes. Prior work in this area has been typified by the use of a single taxonomy for classifying verbal behaviors and limited outcomes (compliance and/or satisfaction). We applied three different IA systems (Bales, Roter’s modified Bales with affective ratings, and Stiles’ “Verbal Response Modes”) to 101 new-patient visits to a general medical clinic for which multiple outcomes had been determined: Several measures of patient knowledge of problems at conclusion of visit; patient compliance with drugs (over the ensuing three months); and patient satisfaction with the visit (perceived technical, interpersonal and communication quality). Within IA systems, cross tabulations and multiple regressions were performed to relate encounter events to outcomes. Across IA systems, multiple regression R2and R2adjusted (R2α) for the number of independent variables entering were used to characterize strength of relationships. Roter’s IA system showed stronger relationships to outcomes of knowledge (41% R2, 27% R2α) and compliance (44% R2, 28% R2α) than did Bales’ or Stiles’ systems. R2for patient satisfaction was identical for Bales and Roter (35%), and greater than R2 for Stiles (14%). We conclude that choice of IA system for research or teaching purposes should be based on behaviors and outcomes of particular interest and importance to the user. Based on audio-review of tapes, Roter’s approach is less time-consuming and may perform as well as more complex systems requiring transcript analysis.

AB - Interactional analysis (IA) systems have been devised and applied to doctor- patient dialogues to describe encounters and to relate process to outcomes. Prior work in this area has been typified by the use of a single taxonomy for classifying verbal behaviors and limited outcomes (compliance and/or satisfaction). We applied three different IA systems (Bales, Roter’s modified Bales with affective ratings, and Stiles’ “Verbal Response Modes”) to 101 new-patient visits to a general medical clinic for which multiple outcomes had been determined: Several measures of patient knowledge of problems at conclusion of visit; patient compliance with drugs (over the ensuing three months); and patient satisfaction with the visit (perceived technical, interpersonal and communication quality). Within IA systems, cross tabulations and multiple regressions were performed to relate encounter events to outcomes. Across IA systems, multiple regression R2and R2adjusted (R2α) for the number of independent variables entering were used to characterize strength of relationships. Roter’s IA system showed stronger relationships to outcomes of knowledge (41% R2, 27% R2α) and compliance (44% R2, 28% R2α) than did Bales’ or Stiles’ systems. R2for patient satisfaction was identical for Bales and Roter (35%), and greater than R2 for Stiles (14%). We conclude that choice of IA system for research or teaching purposes should be based on behaviors and outcomes of particular interest and importance to the user. Based on audio-review of tapes, Roter’s approach is less time-consuming and may perform as well as more complex systems requiring transcript analysis.

UR - http://www.scopus.com/inward/record.url?scp=0020379492&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0020379492&partnerID=8YFLogxK

M3 - Article

VL - 20

SP - 535

EP - 549

JO - Medical Care

JF - Medical Care

SN - 0025-7079

IS - 6

ER -