Table 7 Outcomes related to care processes and cost effectiveness

Ashmore et al., 2014

SCR of 84% (42/50)

Raw agreement,

88% (κ = 0.795;95% CI, 0.58–1.00) between ESP and medical imaging or surgery.

Daker-White et al., 1999 [23]

A greater proportion of PTs ordered no investigations at all (47.5% vs 14.7%; p < 0.001). A smaller proportion of PTs ordered plain x-rays (13.9% vs 41.5%; p < 0.001). No significant differences in other diagnostic tests ordered.

No significant differences in direct costs to the patient or NHS primary care costs.

Direct hospital costs were significantly lower in the PT arm (mean cost per patient £256 vs £498 p < 0.001), as

PTs were less likely to order radiographs and to refer patients for orthopaedic surgery.

Desmeules et al., 2013 [20]

Raw agreement, 88%, between APP and OS (κ = 0.77; 95% CI:0.65–0.88)

No significant differences when ordering any type of imaging tests between APP and OS (p = 0.113).

General inter-rater agreement (κ = 0.65; 95% CI:0.52–0.79), for X-rays only (κ = 0.48; 95% CI:0.33–0.64)

Raw agreement, 88%, (κ = 0.86, 95% CI:0.80–0.93) between APP and OS

Dickens et al., 2003 [70]

Raw agreement among all 3 clinicians: 76.5% (13/17). Correct diagnosis by the orthopaedic consultant 92%, PT 1 84%, PT 2 80%. Diagnostic accuracy for various types of injuries ranged from 96 to 100% for the OS, and from 94 to 98% for the PTs.

Jovic et al., 2019 [68]

SCR 78% with ASP-led service; with prior orthopaedic-led model 38%.

Inter-rater agreement between ASP and OS on treatment (κ = 0.75; 95% CI 0.62–0.89).

Inter-rater agreement between ASP and OS (κ = 0.93; 95% CI 0.87–1.00).

Lowry et al., 2020 [69]

Raw agreement, 70%, on surgery between PT and OS (κ = 0.46; 95% CI 0.21–0.71).

No significant differences in terms of frequency of medical imaging requests were found between Pt and OS; x-rays (p = 0.338), MRI (p = 0.799), other tests (p = 0.400).

Raw agreement, 70%, between PT and OS (κ = 0.42; 95% CI 0.19–0.66).

Raw agreement, 86%, between PT and OS (κ = 0.80; 95% CI 0.67–0.93).

MacKay et al. 2009 [71]

Raw agreement, 85.5% (53/62) on surgery between PT and OS (κ = 0.70)

Raw agreement 69% between PT and OS.

Marks et al., 2016 [67]

Raw agreement, 94%, between PT and OS (AC1 = 0.93; 95% CI 0.90–0.93)

Raw agreement, 88%, between PT and OS (AC1 = 0.87; 95% CI 0.83–0.91)

Raw agreement, 74%, between PT and OS (AC1 = 0.72; 95%CI 0.66–0.78)

Napier et al., 2013 [66]

SCR referred by the APP

91%, vs 22%

of patients referred by a GP or ED physician.

Raw agreement, 84.4% between APP and OS (κ = 0.77; 95% CI, 0.60–0.94).

Oldmeadow et al., 2007 [19]

Raw agreement, 74%, between PT and OS on management decisions (κ = 0.38: 95% CI 0.13–0.63).

Razmjou et al., 2013 [72]

Raw agreement 88% between APP and OS for surgery. The APP tended to recommend surgery more often than the OS, 65% vs 55%, (κ = 0.75; 95% CI 0.62 to 0.88).

Raw agreement 97% on x-rays between APP and OS (κ = 0.91; 95% CI 0.81 to 1.00).

Raw agreement on major diagnostic categories between APP and OS varied from 84 to 98% (κ = 0.68 to 0.94)

Significantly shorter waiting time for APP assessment than for OS assessment at all three time points (Wilcoxon 6.20, 5.92 and 5.41, p < 0.001)

Samsson et al., 2014, 2015, 2016 [24, 58, 57]

Significantly higher SCR with PT triage, 55%, vs standard practice 25%; difference 30% (95% CI 11 to 49%), p = 0.002).

Significantly lower proportion of investigations ordered by the PT (17% vs 29%; difference − 12% (95% CI −23 to 0.6%), p = 0.039.

Significantly fewer days in PT group 19 (SD 12) vs 28 (SD 14) days in the standard practice group (p < 0.001)