Jcehp volume 15 numbers 1,2,3,4

The Journal of Continuing Education in the Health Professions, Volume 15, pp. 31–39. Printed in the U.S.A. Copyright 1995 The Alliance forContinuing Medical Education and the Society of Medical College Directors of Continuing Medical Education. All rights reserved.
Original Article
Patient Charts and Physician Office Management Decisions:
Chart Audit and Chart Stimulated Recall
Professor, Department of Pharmacology & Professor, Office of Medical Education and Faculty of Medicine, The University ofCalgary, Calgary, Alberta Abstract: Accurate assessment of clinical competence and performance in office prac-
tice is enhanced through a multi-tool approach. Two assessment tools that offer a
complementary range of information, specific to the patient’s chart, are chart audit
(CA) and chart stimulated recall (CSR). This paper demonstrates how chart audit and
chart stimulated recall provide insights into the office management of osteoarthritis in
the elderly. CA provides basic data for clinical choices when the areas of problem iden-
tification, history, physical, investigations, and treatments are examined. CSR illuminates
the rationale behind decisions, as well as the choices considered and the options ruled
out. Furthermore, CSR shows how individual patient and physician characteristics,
practice and professional factors, and health care system and social factors, are influ-
ential variables on the physician’s clinical management decisions. Supplementing the
type of data extracted from the CA with those found through CSR allows for a broad
range of information to be used in assessing a physician’s ability to make clinical deci-
sions. Physicians, educators, and assessors, will benefit from considering the value of
using both of these patient chart approaches when reviewing clinical care.
Key Words: Assessment tools, chart audit, chart stimulated recall, clinical compe-
tence, NSAID gastropathy, osteoarthritis, patient chart
With the high rate of osteoarthritis (OA) in the arthritic patients are managed is important.1 elderly, and the potential for drug-related ill- A recent University of Calgary/McGill stu- nesses due to the prescription of nonsteroidal dy examined the diagnostic, investigative, anti-inflammatory drugs (NSAIDs), an under- and treatment decisions for OA geriatric pa- standing of how ambulatory geriatric osteo- tients who visited physicians in their offices Patient Charts and Physician Office Management complaining of stomach pains. Chart audit approached through the Alberta Primary Care (CA) and chart stimulated recall (CSR) were Research Unit, Department of Family Medi- the two techniques used, which utilize the cine, Calgary Branch by means of an estab- lished recruitment process. Thirty-one physi- cians responded to the invitation. Twenty tool for assessing physician clinical compe- physicians participated in the actual study; tence and performance. Traditionally, espe- eleven in the pilot, instrument testing, and cially in the hospital setting, CA has been an training stage. All study instruments were accepted method employed to study clinical piloted, and validity and reliability established, choices.2 In recent years, however, because of the recognized limitations of the CA—par- ticularly in the office setting—the advantageof supplementing the CA with additional tools The physicians were visited in their office by of assessment has been acknowledged.3 A sec- a standardized patient. The practitioners were ond assessment technique, CSR, which also unaware of the patient’s identity and present- uses the patient’s chart, provides additional data ing condition, which was NSAID gastropathy.
that cannot be discerned through CA alone.
Other factors of the patient’s background CSR can determine not only physician choices, included OA of the hip, diagnosed 3 years ear- but the rationale behind those choices. Diag- lier; treatment for a gastric ulcer 10 years prior noses, investigations, and treatment options to the office visit; and controlled hypertension ruled out can also be discovered.3–8 Using both and diabetes. The patient’s prescription med- CA and CSR allows for a broad variance in the ications included hydrochlorothiazide (25 mg once daily), a slow release potassium supple- ment (1 tablet twice daily), methyldopa (250 strate two ways in which patients’ charts can three times a day), glyburide (5 mg twice a day), be used to provide insights into physician decision making in the office. Specifically, needed). The patient was also taking over-the- data collected by CA and CSR provide detail counter ibuprofen (200 mg as needed).
that can assist in understanding the complex- Chart Audit
After the patient’s office visit, either the stan- dardized patient took the patient’s chart to theresearch office in person, or it was mailed in All family physicians (n = 504) in active, from the study doctor’s office. A trained and full-time practice* who saw geriatric patients experienced health records technician com- and practised in the Calgary city area were eli- pleted a CA at the research center. Training had gible to participate. Potential participants were continued until the health record technician’sdata collection findings were consistent with those of the principal investigator—a health record analyst. Data was collected on a stan- further data specific to the presenting problems, dardized CA form, drafted initially by the but did add details as to which conditions principal investigator and tested in a pilot pro- were selected to be primary, and which were ject. Further general details regarding the CA technique can be found in Neufeld, 1985, and Patient’s history and physical exami-
Lockyer, Harrison, and Manning, in press.
nation. Where CA focused on particulars that
were charted by the physician, the CSR tech-
Chart Stimulated Recall
nique elicited additional, important factors; Using the patient’s chart as a stimulus for specifically, in the study of the 20 cases, 5 for recall, the physicians were interviewed after patient’s history and 2 for physical examina- the office visit by a trained nurse using the CSR tion. In particular, these details were related to method.4 The interviewer used a standardized the following factors: physician inquiries made protocol to elicit information specific to the but not recorded; signs present or absent but physician’s management of a patient with not charted; or to an expansion as to why a par- NSAID gastropathy—the rationale and deter- ticular decision was made (e.g., “patient is a minants for clinical choices, the conditions smoker”, or “Tylenol is no longer effective”).
ruled out, and the reasons. Problem identifi- Investigations. While CA was able to pre-
cation, history, physical, treatment, and follow- cisely determine which tests were ordered, up issues were addressed. Interviews lasted the additional CSR data revealed which tests were considered, but not ordered. As well, taped and transcribed. Content analysis of the with CSR, the rationale behind these deci- transcriptions was carried out by a trained research assistant using qualitative method- Treatment. CSR allowed for unrecorded
ology.13 A classification system for coding treatments to be recalled by the physician. For responses into categories was developed and example, although CA was efficient in its abil- patterns and trends were observed. The CSR ity to reveal the prescriptions that were writ- methodology has been described in depth in ten down, it failed to uncover the nonphar- macologic treatments elicited by CSR. Aswell, with CSR, the rationale for decisions were given, as were the treatments that wereconsidered, but not ordered (Table 2).
Clinical Management
Follow-up. CSR allowed for additional
information to be included. Although both CA Problem identification. Chart audit alone
and CSR revealed the time period in which the captured the presenting complaint and, at patients would be recalled, CSR gave the ratio- times, the closely associated differentials. CA nale as to why the physician felt 2 weeks were was also adequate in recognizing the sec- preferable to one, or vice versa. For example, ondary diagnoses—in this case, controlled “[NSAID gastropathy was] a serious condition hypertension and diabetes. CSR did not elicit [that] needs close follow-up”, versus, “.give Patient Charts and Physician Office Management Investigative Details as Determined by Chart Audit and Chart Stimulated Recall
Chart Audit
Chart Stimulated Recall
Investigations
Ordered (%)
Considered (%)
Rationale
time for [the] medications to work”, or for became apparent through CSR. Seven patient test results or previous records to arrive.
factors surfaced as determinants, as well as 26physician characteristics, five practice ele- Context of Clinical Choices
ments, and seven health care system factors.
Several patient characteristics, physician char-acteristics, practice or professional factors, Discussion
and health care system and social factors wereoften stated to be influential variables in clin- Although for some time the medical record, ical decisions (Table 3). Their impact only through CA, has played a central role in Treatment
Chart Audit
Chart Stimulated Recall
Treatment
Ordered (%)
Considered (%)
Rationale
had samples; diagnostic trial, aswell as therapy, confirms the diag-nosis; easier to take, qid, bid, bettercompliance; doesn’t react with othermedications; less side effects, otherthan GI; well tolerated in the elder-ly; gives better/quicker results;doesn’t like to prescribe until sureof diagnosis; familiar; multiple times a day, less compli-ance; reluctant to use with elderlypatients, not well tolerated; cancause other side effects; doesn’t liketo prescribe until sure of diagnosis;reacts with other medications decreases acid but doesn’t restoremucus barrier; doesn’t like to pre-scribe until sure of diagnosis had samples; inexpensive; givesbetter/quicker results multiple times a day, less compli-ance; reestablishes mucus barrierin stomach Patient Charts and Physician Office Management Table 2 (continued)
Treatment
Chart Audit
Chart Stimulated Recall
Treatment
Ordered (%)
Considered (%)
Rationale
hospital quality assurance programs, more agement approaches, identifying areas for recently, insights into how the patient’s office learning, and designing CME. The rationale chart can further be used to understand and behind decisions, and choices considered and ruled out, become part of the picture. As well, demonstrated.3 Findings from this study sup- the practice care context, the health care sys- port this observation. CA, supplemented by tem, patient’s capabilities, and time constraints, CSR, can elicit information about clinical among others, are acknowledged often as rele- vant factors. As other factors can come into choices, and the rationale behind those deci- play when clinical competence is translated sions. It can also permit individual patient, into practise, it is necessary to view education, physician, practice, and health care determi- learning, and assessment more broadly than in CA and CSR, have specific purposes, there learning, and continuing medical education are overlapping areas or items (e.g., what was done on history, physical, investigations, treat- confirm that there are additional factors, ment, and follow-up). At times in our study, dis- other than physician competence, which are crepancies between the findings elicited by significant to consider when assessing man- CA and CSR in these areas were observed.
Factors That Influence Patient Management
Demographic Factors
Factor Present (%)
Rationale
if patient not bright, would need to instruct caregiver if patient not bright, would need to instruct caregiver;if patient is bright, not cognitively impaired, will beless parental/directive if patient is bright, not cognitively impaired, will beless parental/directive Where Physician Learned Approach
Factor Present (%)
Formal TrainingWhere physician was trained Practice Style
First Visit
Doesn’t do a battery of tests on first visit, no improvement, then tests
On first visit deals only/mainly with presenting complaint Uses first visit to get acquainted and build rapport Does an exam and tests on first visit, discusses results and goals on second Second VisitDoes complete physical on second visit Will make extra time, even if busy, if it is difficult for patient to come back for second visit GeneralPatient will do what they want to do, so must work with what patients will do If patient doesn’t return, phones patient and encourages follow-up Must take into account cultural factors, affects patient expectations and compliance Doesn’t refer to specialist unless case is serious or patient insists Patient Charts and Physician Office Management Table 3 (continued)
Factors That Influence Patient Management
General (continued)Doesn’t do a lot of tests if considering consulting because consultant will just redo tests Sees role as an advisor, doesn’t like to take control of patient’s illness Tends to be abrupt, doesn’t discuss non-medical things Decides on whether to do exam and how much time to spend with patient New physician so has more time to spend with patients Resident in practice (practice used for resident training) Does not provide services that he/she doesn’t get paid for Tries not to let economic concerns interfere with services provided and time spent with patient Salaried physician, can spend more time with patients Does not feel provincial health system pays enough to do complete physical exam on first visit Wants laboratory in office building but government will not allow any more labs in doctors’ offices These variations merit comment. Limited chart- understand are the strengths and limitations of ing practices or habits can possibly help explain each tool,2,3,8,12 including costs, when deciding CA findings being lower in some instances upon their use or feasibility. CA and CSR than the CSR. The physician may have been techniques, in combination, should be em- able to recall the details during the CSR inter- ployed only when the purpose, study, or activ- view, although the particulars have not been documented. Alternatively, due to the depen- Conclusion
dence on memory and self-report during theCSR interview, the physician may have artic- Although this study has limitations, it serves ulated something having been done that really to illustrate how the patient’s office chart, had not. Although not the purpose of this arti- specifically a combination of CA and CSR, can cle, it is important to note that if variation help in capturing the factors associated with occurs when applying more than one assess- clinical decisions. Such data offers a comple- ment tool that accurate findings be sorted out.
mentary range of information, which can be Project limitations, such as relatively small useful to professional assessors, educators, sample size and focus on one clinical condi- doctors, and future physicians. The value of tion, are acknowledged. Also important to such observations cannot be underestimated for medical education activities such as cur- determining needs in continuing medicaleducation. Proceedings of the Annual Conference on Research in MedicalEducation, 1987; 26:103–108.
Acknowledgments
7. Jennett PA. The patient chart: a tool in the measurement of clinical reasoning and per-formance. Further developments in assessing James Neary provided technical assistance, Publications, INC Congress Centres, Ottawa, Canada 27–30 June, 1987; 69:680–688.
Foundation gave financial support for this Solomon DG. Designing a reliable and validchart stimulated recall examination. Paperpresented to the Annual Meeting of the References
Society of Academic Emergency Medicine.
Minneapolis, Minnesota, 1990.
1. Neuberger JS, Neuberger GB. Epidemiology 9. Grad RM. Evaluation of a test to assess of rheumatic diseases. Nurs Clin North Am physician knowledge of drug prescribing for the elderly. Abstract. Congress on Quality of 2. Neufeld VR, Norman GR. Assessing clinical Care in Family Medicine/General Practice.
competence. New York: Springer Publishing 10. Jennett PA. “Pilot test for chart audit stimu- 3. Jennett PA. Chart stimulated recall. A tech- lated recall (CSR) interviews: osteoarthritis nique to assess clinical competence and per- (OA)”. Research grant submission to The formance. Educ Gen Pract 1994; in press. 4. Jennett PA, Tambay JM, Atkinson MA, et al.
Chart stimulated recall: a method for assess- 11. Jennett PA. “Standardized patient measure of ing factors which influence physician’s prac- physician management approach, osteoarthri- tis (OA)”. Research grant submission to The assessment of clinical competence. Part 2.
International Conference Proceedings 1992.
12. Lockyer J, Harrison VR (with assistance Scotland. 1–3 September 1992, Abstr. 4C4.
from Philip Manning). Performance assess- ment: the role of chart review, analysis and continuing medical education in physician munication between physician and evaluator performance enhancement. In: Davis D, Fox on assessments of clinical performance.
R, eds. The physician as learner—imbedding 13. Miles MB, Huberman AM. Qualitative data 6. Parboosingh J, Avard D, Lockyer J, Watson analysis. Beverly Hills: Sage Publications, M, Pim C, Yee J. Interviews as a method of

Source: http://sacme.ywd.ca/site/sacme/assets/pdf/JCEHP_award1995.pdf