Table 6.

Summary of clinical prediction rules and classification systems for painful musculoskeletal conditions

Condition	Purpose of rule/system	Evaluation status of the rule/system
Low back pain	Identifying responders to spinal manipulation [77–97, 119, 120]	Rule developed empirically Confirmatory testing in additional samples, including workers, has shown mixed results Unsuccessful evaluation via clinical trial Patients positive for the main factors in the rule (symptom duration <16 days and no symptoms distal to the knee) respond to other interventions such as McKenzie therapy
Low back pain	Treatment-based classification system [55, 56, 97–111]	Rule developed theoretically Empirical testing has had mixed results Mixed evaluations via clinical trial Reliability of mobilization and stabilization subgroups has been unacceptable in some studies Questionable utility of the system as 25 % of patients fall into multiple subgroups and 25 % remain unclassified
Low back pain	Identifying responders to stabilization exercise [92, 97, 110, 112]	Rule developed empirically Confirmatory testing in additional samples has had mixed results Stabilization and manipulation rules do not represent mutually exclusive subgroups
Low back pain	Identifying responders to McKenzie approach [95, 97, 113]	Rule developed empirically No confirmatory testing Prevalence testing of McKenzie classifications shows overlap with other prediction rules
Low back pain	Identifying non-responders to spinal manipulation [114]	Rule developed empirically No confirmatory testing
Low back pain	Identifying responders to mechanical traction [115, 116]	Two separate rules developed empirically Factors in the rules are not consistent No confirmatory testing
Low back pain	CBI health classification system [117]	System developed theoretically Empirically tested by the developers No confirmatory testing
Low back pain	Identifying responders to pilates based exercises [118]	Rule developed empirically No confirmatory testing
Neck pain	Treatment-based classification system for neck pain [123, 124, 128–131, 135, 137]	System developed theoretically Empirical testing has been promising (case series, pilot and cohort studies) No evaluation via clinical trial
Neck pain	Identifying responders to Thoracic manipulation [102, 125–127]	Rule developed empirically Unsuccessful evaluation via clinical trial
Neck pain	Identifying responders to cervical traction and exercise [133, 134]	Rule developed empirically No confirmatory testing
Neck pain	Identifying responders to home-based cervical traction [122]	Rule developed empirically No confirmatory testing
Neck pain	Identifying responders to cervical manipulation [132, 136]	Two separate rules developed empirically Factors in the rules are not consistent No confirmatory testing
Neck pain	Identifying responders to cervical manipulation physiotherapy or usual care [138]	Rule developed empirically No confirmatory testing
Patellofemoral knee pain	Identifying responders to lumbopelvic manipulation [141, 142]	Rule developed empirically Confirmatory testing in a separate sample was unsuccessful
Patellofemoral knee pain	Identifying responders to foot orthosis [140, 146]	Two separate rules developed empirically Factors in the rules are not consistent No confirmatory testing
Patellofemoral knee pain	Identifying responders to patellar taping [143]	Rule developed empirically No confirmatory testing
Ankle sprain	Identifying responders to manipulation and exercises [148]	Rule developed empirically No confirmatory testing
Lateral epicondylalgia	Classification model for tennis elbow [149]	Theoretical model description No empirical testing
Lateral epicondylalgia	Identifying responders to manual therapy and exercise [147]	Rule developed empirically No confirmatory testing
Thoracolumbar injury	Classification system for Thoracolumbar spine injury [144, 145]	System developed theoretically Successful reliability testing No confirmatory testing