Concepts underlying outcome measures in studies of consultations in general practice

Abstract

Objective. To give an overview of the concepts used to describe and evaluate the outcome of general practice consultations. Method. A literature study was undertaken. Among 101 articles relevant to the subject 35 were chosen to illustrate the concepts found. Result. The following concepts were found: disease control, patient compliance, efficacy, symptom relief, enablement, general health, and patient satisfaction. The first three concepts encourage objective measurement while the following four concern the patient's subjective assessment. Methods naturally linked to the respective concepts are briefly described. Discussion. The concepts reflect very diverse aspects of general practice. Some of the concepts come from other domains of research, while others have been developed within general practice. Disease is not exclusively in focus. The experience, competence, and life situation of the patient are as well dealt with as they should be in accordance with a patient-centred profession. However, the concepts spring from the researchers’ ideas about what is important to patients. The patients’ priorities were not sought.

• Consultation
• evaluation
• family practice
• general practitioner
• outcome assessment

General practice is an arena for a great variety of illnesses, diseases, worries, needs, expectations, and demands. Patients consult for anything between birth and death. The possible mission is unbounded. Many have symptoms or concerns that are not possible to classify while others have a mixture of diseases and illnesses, the whole of which is as difficult to grasp as it is difficult to perceive in detail. Furthermore, some diseases or illnesses are not curable – they get worse or get better whatever the doctors’ actions.

Therefore, during his or her career, a GP may ask repeatedly: What do I actually achieve – what are the consequences of my consultations?

Donobedian stated that outcomes such as patient satisfaction, physical disability, or social restoration are difficult to define or measure and to use for evaluation of general practice consultations [1]. Howie stated that assessing quality in general practice is handicapped by the absence of an adequate range of outcome measures. Guidelines often recommend biomedical data for evaluation of treatment effects. For self-limiting conditions, multidimensional problems, or health promotion, such measures are not sufficient [2].

Key Question: What concepts have been used in the literature to describe and evaluate the outcome of consultations in general practice?

Answer: The following concepts were found: disease control, patient compliance, efficacy, symptom relief, enablement, general health, and patient satisfaction.

The purpose of this study was to explore the concepts, and their corresponding methods of monitoring, used to evaluate the outcomes of general practice consultations. The results of the evaluations in themselves were not in focus.

Material and methods

We searched the databases PubMed, Medline, Cinahl, Psychinfo, Embase, and Cochrane using the search words Family Practice/General Practice/Family Medicine combined with outcome/outcomes/outcome assessment/treatment outcome and consultation/physician–patient relations/family physician/general practitioner in different combinations. The searches were limited to publications in English 1970–2005. We achieved approximately 3000 hits.

We skimmed titles and abstracts and excluded articles dealing with other specialities and other focuses such as referrals to hospitals, education, articles focusing on the consultation process, and evaluations of auxiliary staff. The articles included dealt with outcome and evaluation and with evaluation of outcome of general practice consultations.

In the reference lists of these articles we found still others. We mechanically went through four years of five family practice journals. Colleagues and researchers contributed a few articles. In this way we identified 101 articles of interest for our purpose.

The procedure has many similarities to the work described by Greenhalgh [3] who found that literature searches of complex matters cannot rely solely on predefined protocol-driven search strategies.

Of the 101 articles 77 evaluated outcomes and 24 described different outcome measures. In these articles we focused on the concepts, and the methods logically linked to them, used to describe and evaluate the outcome of the consultation.

We grouped together articles that were about the same, or related, concepts. As some concepts were used in many articles it was neither purposeful nor manageable to include all articles. Instead we applied a qualitative approach and chose articles from each group to illustrate the concepts found. The articles chosen were those that best illustrated or brought important dimensions to the concepts found. Finally we used 35 articles for this work: 24 actually measuring outcome and 11 describing outcome measures.

This work started as an article in the Swedish Läkartidningen 2002 [4], which has been updated and revised.

Result

To describe and evaluate outcome of consultations in general practice the following concepts have been used

Objective assessment possible:

• disease control;

• patient compliance;

• efficacy.

Patient's own assessment indispensable:

• symptom relief;

• enablement;

• general health;

• patient satisfaction.

Objective assessment possible

Disease control

A cure of the disease relates to the disease-oriented aspect of general practice. The patient as a person is excluded from the follow-up, either by definition or by intention. The disease is referred to as an entity on its own and, thus, the effect of intervention is monitored as a modification of the course of the disease.

Biochemical markers and physiological functions relate very distinctly to the situation in the body. In the UKDPS studies [5] glucose, Hba1c, and blood pressure were followed up and measured with great exactness. Here, as in a recent study from Sweden, those parameters were looked upon as risk factors, i.e. intermediate outcomes [6]. Diabetic complications are the real outcomes. In other studies the same parameters are regarded as outcomes in themselves, as in a study by Hansen [7].

Major pathological events such as myocardial infarction or stroke, and the hardest endpoint of all, death, are registered in the cardiovascular intervention studies, as was the case in the Swedish STOP hypertension study [8].

The course of the disease can also be monitored through symptoms that are considered to be its definite correlates. This has been done, for example, for tonsillitis [9] and uncomplicated urinary tract infection [10].

Patient compliance

Compliance or adherence, the extent to which the patients follow instructions, advice about lifestyle, or prescriptions given, is another dimension of the outcome of consultations. The concept is not uncontroversial, since it implies the unilateral distribution of knowledge and responsibility on the part of the doctor. Concordance, which recognizes the two subjects within the consultation, has been suggested as an alternative term. However, in outcome studies in general practice, compliance/adherence is what has been looked at. It reflects the rapport between doctor and patient, and it is a crucial intermediate outcome on the road to benefiting from drug treatment or other interventions.

Measuring compliance in a valid way implies methodological difficulties [11]. In Stewart's study on the effects of patient-centredness, pills were counted [12]. The patients’ self-reported adherence to medication has also been used, although this method is considered to be less reliable [12]. Compliance/adherence has also been used concerning advice given during the consultation. In one study patients were interviewed by phone to find out if they had followed the GP's recommendations for making lifestyle changes or fulfilling planned actions [13]. In another study in a questionnaire after the consultation patients were asked if they had attempted to modify their behaviour with regard to smoking, alcohol, the use of a safety belt, diet, exercise, stress, and safe sex [14].

Another way of tracing compliance has been to ask patients about their intention to follow the advice, either immediately [15] or after some weeks [16].

Efficacy

The overall efficacy of general practice and primary care, measured through the spin-off from a single consultation into tests, revisits, referrals, or episodes of hospital care, is a very relevant outcome, not least from the perspective of the healthcare organization. Even to the individual patient, efficacy should be highly desirable, getting things sorted out straight away. Stewart studied the frequency of revisits and referrals in relation to patient-centredness [17], Christensen studied the frequency of revisits in frequent attenders after an intervention [18], and Owens found fewer revisits in irritable bowel syndrome patients if the doctor had been patient-centred [19].

Patient's own assessment indispensable

Symptom relief

Some of the symptoms that are presented to the general practitioner are the expression of diseases, the majority of them self-limiting, while others are the signs of strain or imbalance, and still others just occur [20]. When the symptom is neither alarming nor requires specific treatment it is the actual experience of the patient that matters. The relief from the discomfort, or concern caused by the symptom, is the outcome.

The effect of the consultation on certain symptoms is usually investigated by open questions, with inquiries, or with scales. Outcome of consultations for conditions such as back pain [21], neck pain [22], chest pain, fatigue, headache, and abdominal symptoms has been evaluated in this way [23].

When studying the resolution of symptom concern Stewart used a VAS scale [17].

Enablement

As pointed out, many consultations in general practice are about self-limiting conditions, or conditions that are not affected at all by the doctor's actions, or very little. Some patients come for an assessment. Traditional disease-oriented outcome measures, and symptom monitoring, are of very limited relevance in these situations. The Patient Enablement Instrument, PEI, developed by Howie et al., mirrors how the consultation influences the patient's ability to understand the illness, and to cope with the symptom as well as with life as a whole. As a concept, patient enablement emanates directly from general practice. It is easy to use, and concentrates on patient outcome irrespective of cause [24]. PEI was one of several outcome measures in two studies of patient-centredness [25], [26].

In a Norwegian questionnaire on patients’ experience of the consultation, the Patient Experience Questionnaire (PEQ), the last part bears much resemblance to the PEI [27].

These questionnaires were filled in immediately after the consultation and little is known about enablement in the long run.

General health

Many patients have more complex states, such as multiple diagnoses or recurrent or progressive diseases. The distinct disease and its course are subordinate to what is important in life to the person in question. The relation between general health and the effects of disease treatment is far from absolute.

Instruments for monitoring general health are well established and applied in several outcome studies in general practice. The Medical Outcome Study short form –MOS-sf [28] – with the instrument SF-36 [29], [30] and also in a shorter form SF-12, EuroQol with EQ-5D [31], COOP-Wonca [32], Measure Yourself Medical Outcome (MYMOP) [33], and Nottingham Health Profile, NHP [34] are such instruments.

For example, in Bertakis's study, general health, measured by SF-36, was related to the styles of the doctors under study [30] and Little used the MYMOP scale when studying the effect on patient outcome of patient-centredness and a positive approach by the doctor [25].

These instruments can be used at different intervals after the consultation and thus a change in general health can be observed over time.

Outcome measures related to general health, but outside the pathway of the validated protocols, have been introduced by some researchers. Winefield asked patients to grade their health before and after the consultation as percentages of their normal, perceived health [13]. Kaplan, when studying the effects of physician–patient interaction on the outcome of chronic disease, used four levels of health: excellent; good; fair; poor [35].

Patient satisfaction

Satisfaction is the dominating concept in outcome research in general practice, either alone or in combination with other measures. It is also the most person-oriented concept and does not necessarily have any bearing on the illness/disease.

In fact, it is very unspecific, referring to different aspects of the consultation, such as satisfaction with the doctor, the communication, the staff, the accessibility, or the fulfilment of expectations. In a review from 1988 Hall noted that only 4% of 221 studies related satisfaction to the outcome [36].

Patient satisfaction with the consultation has been regarded as an important measure of the outcome of the consultation per se, and has also proved to be significant for the healing process, and for compliance with the prescriptions or advice given [13].

Questions on satisfaction are often posed immediately after the consultation. When patients report satisfaction immediately, they refer it to the doctor's behaviour and communication, but later on – after two weeks and three months – they refer satisfaction to the outcome of the consultation [37].

Satisfaction has often been measured with different scales. Hall's review found 75% of the 221 questionnaires were home-made or used only once [36]. In the home-made inquiries, the patients expressed 10% higher satisfaction than in the validated inquiries. The number of questions is of importance: the more questions, the less satisfaction. Nowadays validated scales are more commonly used. The Consultation Satisfaction Questionnaire (CSQ) [38], [39] and Medical Interview Satisfaction Scale (MISS) [39] are validated questionnaires to measure patient satisfaction with different aspects of doctors’ performance. MISS was originally elaborated in the USA as MISS-29 but there is also a version adapted to an English standard, MISS-21 [40].

Discussion

Comments on method

The literature search was complicated and did not fit the shape of an ordinary systematic review. We started to collect articles systematically but got so many hits that it was difficult to manage. With growing knowledge of the field we noted that we simply did not find many of our most important references with search methods that used predetermined criteria. Instead we found them with extended search methods described in the methods section. Our experiences are in line with others who have found that a search in the databases does not yield all suitable articles, for example for systematic reviews of complex evidence [3], for family/general practice subjects [41], or RCTs for systematic reviews [42]. As the work proceeded we found that a qualitative approach to the whole material gave a comprehensive and clear answer to our question. Thus, for the result section, we selected articles to exemplify and illustrate the different concepts that we had found through the analysis. We strove for conceptual overview rather than for numerical completeness.

Comments on the results

We found seven main concepts used to evaluate GPs’ consultations: disease control, patient compliance, efficacy, symptom relief, enablement, general health, and patient satisfaction. On the one hand there are variables that lend themselves to objective measurement, namely disease control, efficacy, and patient compliance. On the other, there are variables that address the patient as a subject. These are symptom relief, enablement, general health, and patient satisfaction.

Disease control relates to the physical body and its immediate correlates in the patient's experience. Efficacy is related to the organization and mirrors the diagnostic and therapeutic skills of the profession. Patient compliance is an intermediate outcome dependent on the biomedical treatment rational. Contrasting, symptoms put experience itself in focus, and enablement, patient satisfaction and general health are related to the patient as a person in his or her context.

Thus enablement, patient satisfaction, general health, and, to some extent, efficacy are concepts that can be used for evaluation irrespective of disease. This could be important in general practice where the patients’ problems are so disparate [2]. On the other hand, cure of disease or symptom relief cannot be excluded from evaluation as they are the primary aims when within reach.

As was stated by Donabedian and Howie, outcome measures for general practice are difficult to find. The best defined outcome measure – death – has not been used to evaluate consultations but is discussed as an outcome measure of practices [43].

A comprehensive approach to outcome evaluation has not yet been designed. Is it sufficient to combine the concepts and methods already available, or is there a need for the development of new approaches?

Combinations of instruments in a systematic way may be a possibility for the evaluation of general practice consultations. One system of systematic combination of methods is HRQoL, Health Related Quality of Life [44]. HRQoL has not been applied to evaluate consultations in general practice.

Although one should not expect to find a hidden golden rule that surpasses the concepts and methods applied so far, a lesson from our study is that evaluations are expressed from the outlook of the profession exclusively. This is a paradox for a discipline claiming patient-centredness to be perhaps its core skill. What are the most important outcomes in the view of the patients?

That question still remains as a result of this study, and it demands further inquiry. We do believe that there is more to be learnt about outcomes in general practice.