- timing of surgery (running late vs on time)
- family history (complication of sinusitis vs none)
- housing (good vs poor)
- nutrition (good diet vs poor)
- patient’s frequency of attendance (frequent attendance vs infrequent attendance)
- patient’s belief in antibiotics (positive vs uncertain).
We then constructed a prescribing decision “dilemma” based on the 2 clinical vignettes, both of which contained plausible symptoms of acute sinusitis in line with Social Judgment Theory.20 One case (PATIENT A) was unlikely to be acute bacterial sinusitis. The other case (PATIENT B) was a probable case of acute bacterial sinusitis. We used Berg and Carenfelt’s evidence-based predictive diagnostic criteria to ascertain diagnostic likelihood.
Case: Patient A
It is a routine morning surgery and you are running on time (timing of surgery). Your patient is female, aged 20 years, and you remember that her brother was admitted with meningitis following sinusitis the year before (family history). She lives in a high-rise apartment in the middle of town (housing), and you know she has a poor diet (nutrition). You have seen this patient 3 or 4 times in the past year (patient’s frequency of attendance). She is complaining of discomfort in the face, clear mucus from both nostrils, and other symptoms of a cold, and she generally feels under the weather. She says she is unsure whether antibiotics would help her (patient’s belief in antibiotics). On examination you confirm the discharge and note mild to moderate tenderness over both maxillary sinuses. Her temperature is 37.4°C (99.3°F) and she has been ill for a week (symptoms/signs). (Negative prediction criteria)
Case: Patient B
It is a busy afternoon in surgery and you are running 15 minutes late (timing of surgery). Your patient is female, aged 20 years, and you remember that her brother was admitted with meningitis following sinusitis the year before (family history). She lives in a high-rise apartment in the middle of town (housing), but you know she has a good diet (nutrition). You have not seen this patient over the past 2 years (patient’s frequency of attendance). She is complaining of localised pain over the right cheek, symptoms of a cold, mucopurulent discharge from both nostrils (worse on the right), and she generally feels under the weather. She says she thinks antibiotics would help her (patient’s belief in antibiotics). On examination you confirm the discharge and note moderate tenderness over the right maxillary sinus and less so on the left side. Her temperature is 37.4°C (99.3°F). She had been ill for a week, then felt better before getting worse again over the last day (symptoms/signs). (Positive prediction criteria)
These plausible scenarios exhibit differing levels of diagnostic likelihood and have been enmeshed with the various permutations and combinations of important decision-making factors (in bold).
The doctors were then asked, “What is the likelihood of your prescribing immediately for this patient?” and “What is the likelihood of your prescribing delayed antibiotics for this patient?” Responses were recorded on 6-point Likert scales ranging from highly likely to highly unlikely. Doctors were also asked about their beliefs—ie, how effective they think antibiotics are in treating acute sinusitis (and complications separately) on 5-point Likert scales (4=very effective, 3=effective, 2=indifferent, 1=ineffective, 0=very ineffective).
The final questionnaire included:
- Family doctor details and demographics (age, gender, number of years as a family doctor, number of hours worked per week), and the health centre (type, list size of registered patients, number of partners)
- A clinical features section itemizing symptoms and signs used to make a diagnosis of sinusitis
- The 2 clinical vignettes incorporating clinical and psychosocial variables
- Questions on the likelihood of prescribing, and questions about the beliefs of the effectiveness of antibiotics.
TABLE 1
Symptoms and signs family physicians used most often for diagnosing acute sinusitis (pain criteria are top 4)
| SYMPTOM/SIGN | MEAN SCORE (95% CI) |
|---|---|
| Tender on facial pressure or percussion | 2.99 (3.07–2.91) |
| Pain in face on bending forwards | 2.92 (3.00–2.84) |
| Severity of facial pain | 2.85 (2.93–2.76) |
| Unilateral facial pain | 2.73 (2.82–2.64) |
| Pus exuding from ostium | 2.70 (2.81–2.58) |
| Frontal pain | 2.55 (2.64–2.47) |
| Previous sinusitis | 2.35 (2.43–2.26) |
| Temperature >38°C | 2.29 (2.39–2.18) |
| Pain in teeth | 2.28 (2.38–2.18) |
| Malaise/unwell | 2.27 (2.36–2.17) |
| Facial edema | 2.26 (2.39–2.14) |
| Purulent rhinorrhea more marked on one side | 2.15 (2.25–2.04) |
| Purulent secretions in the nasal cavity on inspection | 2.09 (2.20–1.97) |
| “Double sickening” after previous cold or upper respiratory tract infection | 2.05 (2.15–1.95) |
| Duration of symptoms more than 7 days | 1.97 (2.07–1.86) |
| Purulent rhinorrhea on both sides | 1.61(1.71–1.50) |
| Blocked nose/nasal congestion on one side | 1.26 (1.35–1.17) |
| Poor response to decongestants | 1.25 (1.34–1.16) |
| Post–nasal drip | 1.19 (1.29–1.10) |
| Hyposmia (reduced smell)/anosmia (no smell)/cacosmia (bad smell) | 1.16 (1.26–1.07) |
| Known allergy | 0.91(1.00–0.83) |
| Blocked nose/nasal congestion on both sides | 0.89 (0.97–0.80) |
| Transillumination | 0.84 (0.93–0.75) |
| Nasal voice | 0.79 (0.87–0.71) |
| Rating: 4= very important, 3= important, 2= moderately important, 1=slightly important, 0=not important | |