4J
ORAL ABSTRACTS: PRACTITIONER DECISION MAKING
4J
Method(s): Q-methodology is a research method that seeks to understand subjective experience and beliefs. Subjects are asked to rank statements (the Q-sample) along a quasi-normal grid, in which either extreme represents opposites on a single face-valid dimension (e.g. disagree-agree/unimportant-important) in a process known as a Q-sort. As study participants sort the Q-sample, patterns (shared views of the phenomenon being studied) start to emerge that can be analysed and interpreted using factor analysis.
Result(s): The presentation will be illustrated using examples drawn from an on-going study of nurses working in acute care settings in one Scottish Health Board, and their decision making around prescribing practice. Prescribing is a new area of practice for clinically experienced nurses with additional training within the United Kingdom healthcare environment. Little is known about patterns of prescribing by this group of practitioners in the secondary care setting, but a preliminary study by the author showed variations in practice that could not be explained by clinical caseload. A further study to investigate the cause of these variations is being undertaken; Q-methodology has been used to understand the factors (including intrinsic clinician factors, and extrinsic patient, drug, or environmental factors) that influence prescribing decisions in three clinical scenarios - general prescribing, antimicrobial prescribing for presumed infection, and prescribing for patients in pain. This study will be used to illustrate the development of the Q-sample from a literature based concourse, the ranking of influencing factors in the Q-sort, and the prelimanary analysis and interpretation of these data.
Conclusion(s): Q-methodology is a research approach that allows robust collection and analysis of subjective data from a variety of subjects, including clinicians, patients, and health service manaagers. It has advantages over other methods in the speed and acceptability of the data collection process, and the much smaller sample size required for meaningful interpretation. This study demonstrates its use in understanding the factors that drive previously unobserved and unexplained variations in prescribing decison making and practice in a group of nurses working in acute in-patient care settings.
Method(s): A web survey was promoted among convenience samples of primary care clinicians and residents, medical and physician assistant students, undergraduate students, and patients. Respondents provided judgments from which 4 measures of their treatment threshold probability for sore throat could be calculated: direct statement of threshold, judgment of 4 utilities (u(TN), u(TP),u( FN), and u( FP)), judgment of 2 utility differences (u(TP) - u(FN) and u(TN) - u(FP)), and person tradeoff judgments (number of people experiencing the less severe error that would be equal to one person experiencing the more severe error). Wording variants and question presentation orders were randomized. Survey asked demographics, parenting experience, and clinical experience of strep throat and of bad outcomes associated with misses and with unnecessary antibiotics.
Result(s): 950 started survey and 735 (77.4%) finished. Each threshold method was noisy, with responses ranging from 0 to 100%. Responses were characterized as invalid, doubtful, worrisome, and reasonable. Those who judged only some methods before quitting made more unreasonable responses. Intercorrelations among measures ranged from -0.03 to 0.39 when all responses were included, and from 0.61 to 0.75 when only reasonable responses were considered. Mean directly stated treatment threshold probabilities (0.59) were higher than the mean thresholds calculated from component judgments (each 0.43). Thresholds showed a U trend over medical education, with 3rdyear medical and PA students having the lowest thresholds (method means 0.26 - 0.41), while patients (means 0.45 - 0.61) and practicing clinicians (means 0.42 - 0.62) had the highest thresholds.
Data did not support that medical education makes individuals agree more with each other about a threshold value.
Physicians reported familiarity with the concept of a treatment threshold probability, but few reported explicitly comparing a patient’s disease probability to a treatment threshold.
Conclusion(s): Treatment threshold probability judgments are very noisy. Different methods yield different thresholds. Agreement increases when obvious and suspected mistakes are discarded, suggesting people may have an implicit threshold. Perhaps more explicit instruction and communication regarding recommended treatment thresholds could build on this.
Patient’s satisfaction after hip or knee arthroplasty can be very different from surgeon’s satisfaction or objective clinical and radiologic findings. Up to 20% of patients after knee replacement are not satisfied with the procedure. An experienced arthroplasty surgeon often has a feeling that he can predict the success of planned surgery, based on his assessment of patient’s personality. This study was conducted to examine the ability to predict satisfaction by the treating orthopaedic surgeon and other health care providers.
Method(s):
40 arthroplasty surgeons were asked to what degree they can preoperatively predict their patient satisfaction one year after a hip or a knee arthroplasty. The prediction was done using the Visual Analogue Scale (VAS).
Then, 219 consecutive patients scheduled for arthroplasty (118 hip and 101 knee) completed the study. Each patient filled Pain Visual Analogue Scale and Oxford hip or knee score before and year after the surgery. Assessment of satisfaction (actual and predicted) was again performed using the Visual Analogue Scale. It was predicted by the referring surgeon and the patient during the clinic visit and treating physician, physiotherapist and 2-3 nurses during hospitalization. A year postoperatively the satisfaction was assessed by patient and his surgeon (who also tried to “guess” patient’s satisfaction). All assessors were blinded to others.
Result(s):
Arthroplasty surgeons felt that they can predict postoperative patient satisfaction in 90% for hip and 70% for knee arhroplasty. However, no significant correlation was found between predicted patient satisfaction (by the patient himself, the treating orthopaedic surgeon or other health care provider) and the actual patient satisfaction one year after the procedure.
Conclusion(s):
Arthroplasty surgeons strongly feel that they can predict patient satisfaction when they refer the patient to surgery. This study shows that our ability to predict patient satisfaction based on superficial impression of patient’s personality is very low or non-existent. Medical decisions based on these impressions are misleading and should be avoided.
We investigate the job attributes that:
· medical trainees most value in their future specialism
· medical trainees most associate with a career in General Practice.
Combining these we investigate the crisis in General Practice in the UK which is characterised by high levels of exit accompanied by insufficient recruitment.
Method(s):
We identify 36 job attributes and use BWS to elicit (i) the relative importance of those attributes to trainees, and (ii) the extent to which they (dis)associate the attributes with a career as a GP.
The data are analysed by estimation of heteroscedastic conditional logit and scale adjusted latent class models. Logit models are estimated to identify the factors (BWS importance scores, demographics, attitudes etc) which affect trainees' choice of specialism.
Result(s):
From our sample of over 800 trainees, only 20% reported that GP was there first of specialism. Choice models estimated on the career desirability BWS indicate that a good work-life balance, working as part of a team and having control over where one works things are among the most desirable job traits. The opportunity to manage a clinical service and working in a speciality to which entry is competitive were among the least desirable.
A good work-life balance was more than twice as important as recognition or job security, more than 5 times as important as being involved in research and more than 7 times as important as having a good chance of promotion.
We find no significant correlation between the desirability of job traits and the degree to which they are associated with a career as a GP. Plotting BWS Importance Scores against each other (Figure 1) highlights inconsistencies between what trainees seek in their medical career and their expectation of GP.
Undesirable job attributes which were strongly associated with a career as a GP included working in a community-based role and working alone. Excitement was a strongly desired attribute but was not associated with a GP career.
Conclusion(s):
Choice models allow marginal effects to be derived showing the impact of variations in job attribute importance on the likelihood of applying for General Practice.
The results suggest that substantial changes to the (perceived) attributes of a General Practice role are required to increase the proportion of trainees choosing this specialty.
