Family Practice Vol. 19, No. 4, 350-356
© Oxford University Press 2002
Health Services Research |
Severity of episodes of care assessed by family physicians and patients: The DUSOI/WONCA as an extension of the International Classification of Primary Care (ICPC)
Department of Family Practice, Division of Public Health, Academic Medical Center/University of Amsterdam, Meibergdreef 15, Amsterdam, The Netherlands.
IM Okkes; E-mail: i.m.okkes{at}amc.uva.nl
Okkes IM, Veldhuis M and Lamberts H. Severity of episodes of care assessed by family physicians and patients. The DUSOI/WONCA as an extension of the International Classification of Primary Care (ICPC). Family Practice2002; 19:350–356.
Received 25 June 2001; Revised 20 November 2001; Accepted 11 March 2002.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Objectives. The DUSOI/WONCA is included in the second edition of the International Classification of Primary Care (ICPC-2), as an extension to assess the severity of episodes of care. We studied (i) family physician's (FPs') assessment of three DUSOI/WONCA parameters per episode of care; (ii) how these relate to patient and episode of care characteristics, and to the interventions that occur; and (iii) how FPs' and patients' assessment of severity compare.
Methods. Twelve FPs participated and coded patient and encounter data with ICPC. Also, they answered three DUSOI/WONCA questions, that were also answered (after the consultation) by 300 patients. Odds ratios were calculated for the relationships of the severity elements to patient and episode characteristics, and interventions. The relative agreement between FPs' and patients' ratings of severity was assessed.
Results. In 2033 consultations, 2860 episodes of care were documented, with a subset of 411 with a paired assessment by patient and FP. Patients appeared to be less hindered by symptoms/ complaints than the FPs thought, and less optimistic about the prognosis without care than the FP. Clear relationships existed between the FPs' assessment of severity and the patient, encounter and episode of care characteristics. Substantial agreement existed between FPs' and patients' assessment of severity.
Conclusions. This study confirms the feasibility for FPs routinely to code the separate elements of severity for episodes of care, simultaneously using ICPC and DUSOI/WONCA. The studied elements of severity all provide relevant information: the interventions that occurred all related to them in a logical fashion. The FP–patient agreement on severity is satisfactory, also in the sense that it seems realistic to include these elements of severity as a topic in the communication with the patient.
Keywords. Doctor–patient communication, episodes of care, ICPC, prognosis, severity.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The burden of disease may vary considerably in individual patients; this is true for all patients and also for groups of patients having the same diagnosis. Rather different diagnoses can be based on a set of similar symptoms and complaints, and, in choosing a diagnosis, the GP/family physician (FP) will (also) express the assessed severity of the disease. Mild respiratory symptoms, for example, often result in a symptom diagnosis such as cough, or in a ‘catch all’ description such as upper respiratory infection, at the same time indicating that no or only symptomatic treatment is required. The same symptoms can, however, also result in a much more serious diagnosis such as bronchus carcinoma, with a poor prognosis and/or the need for major therapy. FPs' clinical judgment vis à vis the symptoms, complaints and signs is closely interwoven with their view on the prognosis, and the chosen diagnosis may, consequently, form the start of action or ‘non-action’ based on their idea with regard to the prognosis. Obviously, explicit methods to measure/express these considerations would be helpful to understand this intertexture better.
A good example of such an instrument is the Duke/WONCA Severity of Illness Checklist (DUSOI/ WONCA). Since 1993, the WONCA International Classification Committee (WICC) has participated in developing this instrument for use together with the International Classification of Primary Care (ICPC).1,2 The committee recognized that a method is needed to enable an FP to code not only the core elements of episodes of care, but also the level of severity of each episode.3–6 The DUSOI/WONCA severity coding system uses generic parameters to be applied to any health problem, allowing the assessment of its severity.1,2 It is included in the second edition of ICPC as an extension: the combined use of ICPC and the DUSOI/WONCA would enable an FP to characterize each episode of care and to determine and document its severity, based on the following parameters:
- symptoms during the past week;
- complications during the past week;
- prognosis during the next 6 months if no treatment were to be given; and
- treatability, or the need for treatment and the expected response to treatment by this patient.
In a field trial in 16 countries, 22 FPs performed DUSOI/WONCA ratings on 1191 patients with a total of 2488 health problems in the form of a composite severity score for all episodes of care dealt with at a consultation.4 Wide variations in severity were shown both between and within diagnoses. The study confirmed that the DUSOI/WONCA composite score was feasible and potentially useful in family practice.4–6
Little, however, is as yet known about the nature of this potential usefulness, of the assessment of the separate parameters of severity per episode of care, and how these relate to patient characteristics and the interventions that occur. This would seem an important object for study, especially because of the expected mutual dependence of the choice of a diagnosis and the parameters of severity. Also, it is not known how the FPs' and the patients' assessment of an episode's severity compare. In this article, three questions are addressed.
- How do FPs assess the separate elements of severity during routine encounters for episodes of care structured with ICPC?
- How do these ratings relate to patient characteristics and the interventions that occur?
- To what extent do FPs and patients agree on the assessment of severity?
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Data collection
The study took place between October 1994 and December 1995. Twelve FPs participated, all of them used to or familiar with coding with ICPC.7–9 The data collection was part of a study into defensive behaviour of FPs, in which severity data were included.10,11 The FPs collected data during all face-to-face encounters for 2 weeks or until 200 episodes of care were included. At the time the study took place in a practice, one of the authors (MV) was available continuously in order to interview patients with encounters flagged by the FP for the occurrence of defensive behaviour; two control patients, matched for sex and age, were interviewed as well. Patient interviews took place immediately after the consultation. At the end of each consultation, FPs filled out a form with the following variables:
- Patient characteristics: sex, age and date of registering with the practice;
- Episode data: for up to three episodes dealt with, the patient's reason(s) for encounter, the diagnosis(es) and the interventions were coded. Patients' reasons for encounter could be a symptom/complaint (e.g. ‘headache’), a diagnostic title (e.g. ‘migraine’) and/or a request for an intervention (e.g. ‘I need a repeat of my hypertension medication’). To each diagnostic code, the FP added whether it was a ‘certain’ diagnosis (i.e. whether or not the inclusion criteria of ICHPPC-2-Defined were met12), and whether it was the start of a new episode of care or a follow-up encounter.
- Severity: the Dutch translation of the DUSOI used in this study preceded the final version included in the DUSOI/WONCA.1 In this study, the questions on symptoms and complications were collapsed into one question; in the questions about the prognosis (with and without treatment), the term ‘care’ was used instead of ‘treatment’. For each question, five answers were presented. Since it is well known that family practice patients may consult for reasons not related to a health problem to which ‘severity’ would apply (e.g. preventive medication, contraception, advice on child care), ‘not applicable’ was also included. In addition, for the questions on ‘prognosis', FP and patient could indicate ‘don't know’. This resulted in the following interview (the phrasing when a patient was addressed is in parentheses):
- To what extent is the patient (are you) hindered by the symptoms of the episode (this problem)?
not applicable/not at all/slightly/clearly/very much
- What would be the prognosis of this problem without your (your FP's) care?
not applicable/don't know/better/same/worse
- What is the prognosis of this problem with your (your FP's) care?
not applicable/don't know/better/same/worse
- To what extent is the patient (are you) hindered by the symptoms of the episode (this problem)?
Analysis
The selected characteristics of patients and of episodes of care were calculated as percentages for all episodes and, separately, for the subset of episodes with a patient interview.
Odds ratios [95% confidence interval (CI)] were calculated for the relationships of the three severity elements with episode and patient characteristics. For question 1, the answers ‘not applicable’ and ‘not at all’ were collapsed, since they, in fact, overlap: one cannot be hindered by a problem that does not produce symptoms or complaints.
The relative agreement between FPs' and patients' ratings of the three elements of the severity of episodes of care was assessed. In the analysis of the data on prognosis with and without care, only episode data were included in which FP and patient expressed an opinion (i.e. did not answer not applicable or don't know).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In 2033 consultations, 2860 episodes of care were documented. Of these, 4.9% were flagged for defensive behaviour; the 104 patients concerned and 200 controls were interviewed. This resulted in a subset of 411 episodes of care with a paired assessment of severity by patient and FP.
Data on patients and episodes (Table 1
) indicated a predominance of women, and of patients listed in the practice for >8 years. Approximately half of the episodes started during the encounter (‘new’ episodes), and the diagnosis was uncertain in 12.6% of all episodes. One in three episode titles was a symptom/complaint diagnosis. The distribution of the selected interventions, and of reasons for encounter in the form of a request for an intervention, was similar to that in the reference database of the Transition Project.13
|
According to the FPs, in half of the episodes, patients were clearly or very much hindered by their symptoms (Table 2
); while they were generally optimistic about the prognosis without care, they valued the prognosis with care as considerably better. Patients appeared to be slightly less hindered by the symptoms and complaints than their FPs thought they were, and they were less optimistic about the prognosis without care. Patients more often than FPs indicated that they did not know the prognosis with care; however, they appeared to have relatively a lot faith in the effect of care: almost 40% felt that their problem would become worse without it (as compared with 21% of the FPs). With FP care, both FPs and patients almost never expected the problem to get worse (1–2%).
|
In Figure 1
, the FPs' assessment of severity is presented by patients' sex and standard age groups. The FPs' prognosis with and without care was considered worse in older age groups; patients from 25 to 65 years were thought to be hindered most by their symptoms.
|
In Tables 3 and 4
, odds ratios are presented for the three elements of severity as assessed by the FP and 11 characteristics of the episode encounter. When the FP considered a patient to be clearly or very much hindered, more prescriptions were issued, more imaging was ordered and more referrals to a specialist occurred (Table 3). In these cases, relatively often, the diagnosis was a symptom diagnosis, and the encounter a follow-up; also, patients relatively rarely asked for an intervention (Table 4).
|
|
A bad prognosis (‘the same’ and ‘worse’) coincided with more prescriptions, blood tests and referrals to a specialist, and it was linked more often to a disease diagnosis, to follow-up and to a request for an intervention by the patient. In the small proportion of cases where an assessment of the prognosis was not applicable or not known, more tests were ordered and relatively many referrals took place (Table 3
). This was linked relatively often to an uncertain diagnosis, to a follow-up encounter and to a request for an intervention by the patient (Table 4). A less optimistic perspective on the prognosis with care was linked to relatively few prescriptions and more blood tests, and occurred relatively often during follow-up, in a certain diagnosis. In a symptom diagnosis, linked to a request for an intervention by the patient, relatively few prescriptions were issued. In cases where an assessment was not applicable or unknown, the diagnosis was often uncertain or a symptom diagnosis, and the interventions followed the same pattern as in cases where a less optimistic assessment existed.
In Table 5, agreement between the FPs' and the patients' assessment of severity is presented. Substantial agreement existed in all three elements, and especially in the assessment of the prognosis with care.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This study's data originate from a study into defensive behaviour, and one could wonder whether this affected our results. A comparison of data of defensive episodes with data of the other episodes in fact indicated several small differences: in ‘defensive’ episodes, the agreement between FPs and patients was slightly less in that patients reported more often being hindered by symptoms/ complaints, and having a less favourable view of the prognosis without treatment. Since defensive episodes constituted one-third of the paired assessment cases (the other two-thirds being, in that sense, controls) and 5% of the total sample, the presented results on agreement would, if anything, form a slight underestimation of the ‘real agreement’.
This study seems to confirm that it is feasible for FPs routinely to code the separate elements of severity for each episode of care, using ICPC and DUSOI/WONCA simultaneously.14,15 In order to support this statement further, we looked into the association between the FPs' scores on the three severity measures, since a high association might indicate redundancy. Based on cross tables, we calculated correlations, excluding the category ‘not applicable’. As expected, all chi-squares and all correlations were significant; the correlation between the first and second questions was 0.13, and between the first and third question 0.05. However, the correlation between questions 2 and 3 was 0.47. The cross tables indicated the main reason for this to be that if an FP stated that they did not know the episode's prognosis without treatment, this was true for the prognosis with treatment as well. In a new analysis, excluding ‘don't know’ from both variables, the correlation was 0.21. All in all, there seems to be little association.
The three studied elements of severity all provide relevant information: the interventions that occur all relate to them in a logical fashion. A distinct difference is found between episodes of care with very little or with considerable hindrance by symptoms, with a good prognosis without treatment, and a good prognosis with treatment. In particular, new episodes with a symptom diagnosis are expected to benefit from treatment; on the other hand, a bad prognosis with treatment coincides relatively often with follow-up. Certainty about the diagnosis occurs relatively more often when the prognosis is not expected to improve with treatment. Relatively few prescriptions are given when the prognosis with treatment is not expected to improve. Patients do request interventions relatively often when the prognosis without treatment is considered not very good. These findings were, from a general practice perspective, to be expected, since one would anticipate that the diagnosis and the related interventions are not independent of the burden of disease experienced by the patient, and the prognosis with and without treatment.
This study also shows that the three severity elements give explicit information for the assessment of the utility of the FP's care implicitly included in the diagnosis. The majority of family practice patients are—according to their statement after the consultation—hardly hindered by their health problem, and only see a limited need for treatment, because they think that without treatment a substantial number of problems will improve, and the majority will not get worse. However, once it is decided to give treatment, FPs and patients often agree that this warrants an optimistic view on its utility.
Evidently, inter-doctor variation will result in differences in the assessment of the prognosis of specific diseases with and without treatment. The DUSOI/ WONCA does not presume that every FP using it would or should have the same assessment of the prognosis of, for example, essential hypertension, with or without treatment in one and the same patient, or in all patients with the same diagnosis. Also, it is unlikely that all FPs would have the same estimation of the extent to which a patient is hindered by, for example, a boil on the neck. A substantial proportion of the content of family practice is constituted during the individual encounter between patient and FP, on the basis of their beliefs, on what is said and what is not, and on the negotiations concerning the best strategy to help the patient. Despite ‘rule utilitarian’ guidelines for professional behaviour in specific diseases such as hypertension and diabetes, variation will always occur, also as a result of ‘act utilitarian’ considerations in the contact of this particular patient with this particular FP, at this particular time.19
The overall agreement between FPs and patients on the studied explicit aspects of the clinical approach to a variety of health problems is satisfactory.16–19 It is obviously quite feasible for FPs to discuss these aspects of the burden of disease with their patients. Evidently, it would not be realistic to expect patients always to share their FP's assessment of prognosis and treatability, because often uncertainty about this is exactly the patient's reason for visiting the doctor. Also, it is gratifying that many patients are realistic about the prognosis of their condition and its treatability. This would seem an indication that patients and FPs communicate effectively and have, in general, little difficulty in understanding their respective perspectives on the severity of episodes of care. In a wider perspective, this study provides food for thought in the discussion as to what extent evidence-based family practice should be based on an objective improvement in the patient's condition that could not have been reached otherwise, or on the utility of the doctor–patient interaction in dealing with the patient's worries and the patient's sick role.19,20
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 WONCA International Classification Committee. ICPC-2. International Classification of Primary Care, 2nd edn. Oxford: Oxford University Press, 1998.
2 Okkes IM, Jamoulle M, Lamberts H, Bentzen N. ICPC-2-E. The electronic version of ICPC-2. Differences from the printed version and the consequences. Fam Pract 2000; 17: 101–106.
3 Parkerson GR Jr, Bridges-Webb C, Gervas J et al. Classification of severity of health problems in family/general practice: an international field trial. Fam Pract 1996; 13: 303–309.
4 Parkerson GR Jr, Broadhead WE, Tse C-KJ. Health status and severity of illness as predictors of outcomes in primary care. Med Care 1995; 33: 53–66.[ISI][Medline]
5 Parkerson GR Jr, Hammond WE, Yarnall KS. Feasibility and potential clinical usefulness of a computerized severity of illness measure. Arch Fam Med 1994; 3: 968–974.[Abstract]
6 Parkerson GR Jr, Broadhead WE, Tse C-KJ. The DUKE severity of illness checklist (DUSOI) for measurement of severity and comorbidity. J Clin Epidemiol 1993; 46: 379–393.[ISI][Medline]
7 Lamberts H, Okkes IM. Episodes of care: a core concept in family practice. J Fam Pract 1996; 42: 161–167.[ISI][Medline]
8 Hofmans-Okkes IM, Lamberts H. The International Classification of Primary Care (ICPC): new applications in research and computer based patient records in family practice. Fam Pract 1996; 13: 294–302.
9 Van Boven K, Dijksterhuis P, Lamberts H. Defensive testing in Dutch family practice. Is the grass greener on the other side of the ocean? J Fam Pract 1997; 44: 468–472.[ISI][Medline]
10 Veldhuis M. Defensive behavior of Dutch family physicians. Widening the concept. Fam Med 1994; 26: 27–29.[Medline]
11 Veldhuis M, Wigersma L, Okkes I. Deliberate departures from good general practice: a study of motives among Dutch general practitioners. Br J Gen Pract 1998; 48: 183–186.
12 ICHPPC-2-Defined. International Classification of Health Problems in Primary Care, 3rd edn. Oxford: Oxford University Press, 1983.
13 Okkes IM, Oskam SK, Lamberts H. Van klacht naar diagnose. Epsiodegegevens uit de huisartspraktijk [From complaint to diagnosis. Episode data from family practice]. With CD-ROM. Bussum: Coutinho, 1998.
14 Eccles M, Steen N, Hutchinson A, Bradshaw C, McColl E. Severity measurement using a generic instrument. A feasibility study in ambulatory care involving patients with diabetes or asthma. Eur J Publ Health 1997; 7: 205–209.
15 Shiels C, Eccles M, Hutchinson A, Gardiner E, Smoljanovic L. The inter-rater reliability of a generic measure of severity of illness. Fam Pract 1997; 14: 466–471.
16 Hofmans-Okkes IM. An international study into the concept and validity of the Reason for Encounter. In Lamberts H, Wood M, Hofmans-Okkes IM (eds). The International Classification of Primary Care in the European Community. Oxford: Oxford University Press, 1993: 34–42.
17 Ringmann Fagenberg C, Kragstrup J, Stovring H, Rasmussen NK. How well do patient and general practitioner agree about the content of consultations? Scand J Prim Health Care 1998; 17: 149–152.
18 Undén A-L, Elofsson S. Health from the patient's point of view. How does it relate to the physician's judgement? Fam Pract 2001; 18: 174–180.
19 Lamberts H, Hofmans-Okkes I. Values and roles in primary care. J Fam Pract 1996; 42: 178–180.[ISI][Medline]
20 Jung HP. What makes a good general practitioner: do patients and doctors have different views? Br J Gen Pract 1997; 47: 805–809.[ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  N. Bentzen Family Medicine Research: Implications for Wonca Ann. Fam. Med, May 1, 2004; 2(suppl_2): S45 - S49. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||