Table 1
Author, date and countryPatient groupStudy type (level of evidence)OutcomesKey resultsStudy weaknesses
Rogers SN et al, 1995, UKAll patients referred for facial radiographs over a month. 60 of 65 standard OM views taken during one month, viewed by maxillofacial doctors, emergency doctors and radiology staffRetrospective cohort.Missed fracturesOne facial fracture was missed by the radiology consultant, 1/2 maxillofacial doctors, 3/3 emergency doctors and 2/2 PRHOsSmall number of radiographs with only 7 fractures total. In 6 of 65 cases, the gold standard used remained undecided whether a fracture existed
Sidebottom AJ et al, 1996, UK137 patients referred for facial radiographs between Nov 1994 and Apr 1995. Emergency department SHOs opinionProspective cohortOM15 view onlySensitivity 87.5% Sensitivity 83%Inherent bias in the study design. The same doctor was relied upon to truthfully comment on the OM15 view before looking at the two additional views. Blinding questionable. Insufficient information given as to how sensitivity and specificity worked out—calculations from the same data do not agree
OM15 OM30 and lateral viewSensitivity 87.5% Specificity 97%
Raby N and Moore D, 1998, UKFacial radiographs of 50 patients with a facial fracture and 50 without. Films viewed by 3 radiology doctors, with and without the lateral viewsRetrospective cohortSensitivity for diagnosis of facial fractureSensitivity with and without lateral view remained 90%Radiology doctors used in study, not emergency doctors. Gold standard not described
Sidebottom AJ and Lord TC, 1998, UKAll patients referred for facial radiographs over a year. All patients had only one OM15 viewProspective cohortPatient referral to maxillofacial surgeon130 referrals, 36 had midfacial fracturesNo gold standard used—if a fracture was not spotted on single film, it would have been missed by the study This makes for a fundamentally flawed study
Number of maxillofacial referrals the pervious 12 months131 referrals. Number of fractures unclear
McGhee A and Guse J, 2000, UKSelection of facial radiographs for 44 patients with a fracture, and 49 patients without a fracture Emergency doctors asked to reportRetrospective cohortClinical utility for detection of fracturesSame cohort of doctors reported all three combinations of radiographs, which introduces bias
OM15 films onlySensitivity 89.4% Specificity 82.1%%
OM 30 films onlySensitivity 88.6% Specificity 84.8%
Both films togetherSensitivity 90.9% Specificity 94.8%
No statistical significance between values