Table 4.
Methodical quality of the included studies
| Burri et al. | Charles et al. | Frese et al. | Nielsen et al. (2001) | Nielsen et al. (2004) | Okkes et al. | |
|---|---|---|---|---|---|---|
| Domain A: Selection of patients and GPs (refers to all studies regardless the review question) | ||||||
| Was the symptom to be investigated clearly described? | no | no | no | no | no | no |
| Were the selection criteria of the patients clearly described? | yes | yes | yes | unclear | unclear | yes |
| Was a consecutive or random sample of patients enrolled? | yes | yes | yes | yes | yes | yes |
| Was it a multi-centre study? | yes | yes | yes | yes | yes | yes |
| Did the selection criteria of the patients permit the study population to represent the full spectrum of those presenting with the symptom in the respective setting/ addressed in the review question? | yes | unclear | unclear | yes | no | unclear |
| Were the participating health care professionals/ institutions representative for setting to be investigated in the review? | yes | yes | yes | yes | yes | yes |
| Concern that the selection of patients and GPs introduced substantial variation | low | low | low | unclear | unclear | low |
| Risk that the selection of patients introduced bias: low, unclear, high | low | unclear | unclear | unclear | unclear | unclear |
| Domain B: Data collection and patient flow (refers to all studies regardless of the review question) | ||||||
| Were data about the symptom und the inclusion criteria collected directly from the patients (as opposed to a proxy like a register, routine documentation)? | yes | yes | yes | yes | yes | yes |
| Was the same mode of data collection used for all patients? | unclear | yes | yes | yes | yes | yes |
| Was the number of non-responders/ dropouts unlikely to affect the results? | yes | unclear | unclear | yes | yes | unclear |
| Risk that the mode of data collection and/ or patient flow introduced bias: low, unclear, high | low | low | low | low | low | low |
| Domain C: Determination of the underlying aetiology of the symptom (refers only to review question “What are the underlying conditions and their respective frequencies (differential diagnosis)?”) | ||||||
| Was the etiologic category clearly defined? | Yes (3/6) No (2/6) a (1/6) | Yes (5/7) No (1/7) a (1/7) | Yes (9/10) a (1/10) | Yes (4/6) No (1/6) a (1/6) | Yes (4/6) Unclear (1/6) a (1/6) | Yes (11/12) a (1/12) |
| Was the diagnostic work up likely to correctly classify the respective aetiology? | Yes (5/6) a (1/6) | Yes (6/7) a (1/7) | Yes (9/10) a (1/10) | Yes (3/6) Unclear (2/6) a (1/6) | Yes (4/6) Unclear (1/6) a (1/6) | Yes (11/12) a (1/12) |
| Did every patient receive the same diagnostic work up to detect the respective aetiology? | No (5/6) a (1/6) | No (6/7) a (1/7) | Yes (9/10) a (1/10) | Yes (3/6) No (1/6) Unclear (1/6) a (1/6) | Yes (3/6) No (2/6) a (1/6) | No (11/12) a (1/12) |
| Risk that the diagnostic work up introduce bias | low (2/6) | high (6/7) | high (9/10) | low (3/6) | low (4/6) | High (11/12) |
| unclear (2/6) | a (1/7) | a (1/10) | Unclear (2/6) | unclear (1/6) | a (1/12) | |
| a (1/6) | a (1/6) | a (1/6) | ||||
| Domain D: Determination of the prognosis (refers only to review question “What is the prognosis of patients with the respective symptom presenting in the respective setting?”) | ||||||
| Was the prognostic outcome clearly defined? | Yes (3/3) | - | - | Yes (2/2) | - | - |
| Did the study design include a comparison group without the symptom? | No (3/3) | - | - | No (2/2) | - | - |
| Was the work up/ measurement likely to correctly classify the respective prognostic outcome? | Yes (3/3) | - | - | Yes (2/2) | - | - |
| Did every patient receive the same work up/ mode of data collection to verify the respective prognostic outcome? | Yes (3/3) | - | - | Yes (2/2) | - | - |
| Risk that the prognostic work up introduce bias | High (3/3) | - | - | High (2/2) | - | - |
aThe diagnostic category “no diagnosis” were not judged