Family Practice
© Oxford University Press 1996
Vol. 13, No 5
Printed in Great Britain
Improving patient-doctor concordance:
an intervention study in general practice
ST Liaw, D Young and S Farish
Liaw ST, Young D and Farish S. Improving patient-doctor concordance: an intervention
study in general practice. Family Practice 1996; 13: 427-431.
Objective. This study aimed to examine if providing feedback to the doctor can improve
patient-doctor concordance (PDC) on health problems and treatments.
Method. The study was carried out in a hospital-based primary care service in a lower
socioeconomic status (SES) region of metropolitan Melbourne, Australia. A summary of
the existing patient-doctor concordance on health problems and treatments was presented
to doctors along with a questionnaire seeking their perceptions of and suggestions on how
to act on the findings. In a pre- and post-intervention study, data were collected from con-
secutive new patients who completed a pre- and post-consultation questionnaire seeking
information on the presenting complaint patient-reported health problem, doctor-recorded
health problem, treatments received, and patient expectations of and satisfaction with care.
Diagnostic data were classified into body systems. Descriptive statistics were obtained
and PDC measured. Following the intervention, data collection was repeated to detect any
changes in PDC and patient satisfaction.
Results. The pre-intervention sample (n = 197) was young (mean age 33 years), evenly
divided into English-speaking (48%) and non-English-speaking (52%), and low SES (66%).
The post-intervention samples (n = 95) was similar except for a lower proportion of per-
sons from a low SES (27%). Main body systems reported were musculoskeletal, skin,
respiratory, digestive, urological and gynaecological. Post-intervention, PDC on health
problems improved significantly from 31% to 63% at the problem level (P = 0.001) and
from 65% to 79% at the body system level (P = 0.02). PDC on treatments received also
improved significantly from 5.5 to 6 out of 7 treatment options (P = 0.003). There were
no significant differences due to gender, SES and non-English-speaking background status.
Conclusion. PDC is a practical, useful and relevant indicator of effective patient-doctor com-
munication. A well-presented summary of existing levels of PDC is an effective interven-
tion to improve PDC and, by inference, patient-doctor communication on health problems
and treatments. PDC should also be examined and reported in prevalence and incidence
studies based on patient's reports and doctor's records.
Keywords. Concordance, agreement, communication, records, primary health care.
Introduction
Concordance, according to the Oxford English Dic-
tionary (1990), means "a state of agreement and har-
mony between persons or things"; it does not imply
being right or accurate. Patient-doctor concordance
(PDC) is important from both clinical and research
perspectives. PDC may be operationally defined as the
extent to which two processes will yield the same or
Received 27 February 1996; Accepted 29 March 1996.
General Practice Unit, Department of Public Health and Community
Medicine, The University of Melbourne, 200 Berkeley Street,
Carlton, Victoria 3053, Australia.
a similar result, e.g. will the patient report and the doc-
tor record the same health problems and treatments after
an encounter?
With categorical data, concordance may be usefully
measured by the kappa statistic.1 Sensitivity and
specificity, which require a gold standard, have been
used as 'conditional' indexes of concordance.2 Inter-
observer agreement and variability had been used to
validate patient recall of medications.3 Coefficients of
concordance were used to assess observer variability
generally4 and concordance between mothers' reports
of childhood vaccinations and infections and their
general practitioners' records.3 In addition to a lack of
427

L
428 Family Practice—an international journal
agreement on a consistent statistical measure, concor-
dance research is also hindered by inconsistent
terminology and study designs such as comparing pre-
consultation patient demand for care and medical
records information.
A lack of concordance on medical terminology be-
tween patients and doctors and between doctors has been
recognised for many years.6 Subsequent studies have
demonstrated that:
Patient-doctor concordance (PDC)
• medical records are an inaccurate index of outpatient
drug therapy7 and the true level of depression;8
• the psychosocial content tends to be under-recorded
in the clinical encounter;9
• there are large variations in the agreement between
medical records and the recall of chronic illness,
diagnostic radiography, medication use and reproduc-
tive history;10
• both patient-reported information and medical
records are inadequate for efficient evaluation of
cervical cancer control programs;11 and
• there was variable but generally low PDC on health
problems, especially undefined arthritis and
migraine, and medications, especially over-the-
counter medications and benzodiazepines.12
On the other hand, studies have shown good con-
cordance between the medical record and patient
recollection of important indices of patient-doctor
communications such as chief complaint, name and
number of drugs prescribed13 and the reason for
encounter.14
A clinical approach, where concordance was based
on the identification of the organ system or psychosocial
dimension of the presenting problems, demonstrated that
PDC was worst with psychosocial problems.13 It has
been suggested that patient recall was better when the
disease had clear and unambiguous diagnostic
criteria.16 PDC on the nature of the problem and the
patient's health beliefs had been shown to promote
recovery from acute problems like upper respiratory
tract infection17 and to result in "less dysfunctional
consultations",1* improved compliance,19 and patient
and doctor-reported satisfaction.20
A theoretical framework, encompassing both the
clinical and methodological implications of PDC, should
include patient and doctor factors as well as the health
problem and care involved (Figure 1). PDC is poten-
tially an important indicator of the adequacy and
effectiveness of information sharing among patients
and doctors. Adequate sharing of information about
health problems and treatments ("explaining things")
may well be the most important factor in determin-
ing the quality and continuity of care. This study aimed
to:
(i) develop a practical measure of PDC; and
FIGURE 1 Framework to study patient-doctor concordance
(ii) examine if a simple intervention—providing a sum-
mary of the existing levels of PDC in their practice
to doctors and obtaining their feedback—can im-
prove the PDC over time.
Method
The setting of the study was a hospital-based primary
health care clinic in the northern region of Melbourne
where there was a higher than usual proportion of per-
sons from a non-English-speaking background (NESB).
This service serves a high proportion of persons with
a low socioeconomic status (SES) and from an NESB;
both of these population groups are usually associated
with low health status and special needs.
A multilingual questionnaire (Greek, Macedonian,
Italian, Chinese, Vietnamese, Arabic) was administered
by trained bilingual research assistants, over a period
of 3 weeks, to collect information from consecutive new
patients during routine consultation hours only. Inter-
language and inter-cultural consistency of the instru-
ment was validated by back-translation into English.
Prior to their consultation, patients were asked why and
for what health problem(s) they had come to this
hospital-based primary care facility. Following the con-
sultation, they were asked what they and their doctor
had agreed were their main health problem, treatments,
investigations ordered, their satisfaction with the con-
sultation and whether their expectations were met. The
patient-reported health problem after the consultation,
equivalent to the reason for encounter (RFE) agreed
upon by both patient and doctor, may be a health check,
administrative request, symptom, sign, or a diagnostic
label. The GPs' notes were audited for the main doctor-
recorded health problems and processes of care. The
International Classification of Primary Care (ICPQ was
used to code the health problems and collapse them
into body systems (equivalent to ICPC chapters). EPI-
INFO and SAS were used to analyse the data and
calculate the PDC on health problem and body system