Table 3

GRADE quality assessment of the body of evidence

OutcomeNumber of studiesStudy designROBInconsistencyIndirectnessImprecisionOther considerationsQuality
True positive4 studiesCohortVery serious*Not seriousSerious†Not seriousPublication bias‡
Strong association§
Embedded Image
True negative3 studiesCohortVery serious*Not seriousNot seriousNot seriousPublication bias‡
Strong association§
Embedded Image
False positive4 studiesCohortVery serious*Not seriousSerious†Not seriousPublication bias‡
Strong association§
Embedded Image
False negative3 studiesCohortVery serious*Not seriousSerious¶Not seriousPublication bias‡
Strong association§
Embedded Image
  • *Refers to the ROB within studies.

  • †50% of included studies did not completely adhere to the interventions of interest (lack of SPECT) to answer the research question, hence risking the external validity of findings.23

  • ‡Publication bias could not be fully excluded, but it was not deemed sufficient to downgrade the overall quality of evidence either as the search strategy was extensive and up to date overall.24

  • §Quality of evidence was rated up for magnitude of effect as indirect evidence has shown that early diagnosis increases the probability of full bony healing, which may result in shorter rehabilitation period,5 but mainly for the lack of exposure to ionising radiation as opposed to other modalities.25

  • ¶False negatives present the uncertainty linking to patient-important outcomes, for example, the possible deleterious effects of delayed diagnosis.22

  • ROB, risk of bias.