Skip to main content

Table 2 Factors associated with high-dose opioids reported by individual studies

From: Factors associated with the prescribing of high-dose opioids in primary care: a systematic review and meta-analysis

Study ID [Ref]VariableHigh-doseLow-doseRR (95% CI)NNTH (95% CI)
Count (%)Count (%)
Sociodemographic characteristics
 Chang, 2018a [30]State of residence    
 California47,446 (31%)1,416,000 (36%)0.87 (0.86, 0.87)NA
 Florida54,338 (36%)1,207,982 (31%)1.16 (1.15, 1.17)
 Georgia20,692 (14%)689,886 (18%)0.77 (0.76, 0.78)
 Maryland12,487 (8%)250,868 (6%)1.29 (1.26, 1.31)
 Washington15,866 (11%)330,335 (8%)1.24 (1.22, 1.26)
 Kobus, 2012 [32]Insurance coverage    
 Medicare154 (34%)1352 (28%)1.21 (1.06 to 1.39)NA
Ethnicity    
 Unknown/declined to answer64 (14%)879 (18%)0.77 (0.61, 0.98)NA
Treatment-related factors
 Campbell, 2015 [34]Antidepressants246 (58%)323 (49%)1.18 (1.06 to 1.32)NA
Type of opioid drug    
 Morphine86 (20%)75 (11%)1.78 (1.34 to 2.37)NA
ICD-10 lifetime pharmaceutical opioid dependence49 (12%)28 (4%)2.72 (1.7 to 4.25)NA
ICD-10 12-month pharmaceutical opioid dependence26 (6%)13 (2%)3.11 (1.61 to 5.98)NA
Prescribed opioid difficulty scale (PODS) intermediate-high (≥ 8)297 (70%)367 (56%)1.26 (1.15 to 1.38)NA
Past 3-month tampering38 (9%)29 (4%)2.03 (1.27 to 3.25)22 (13 to 64)
Past 3-month different drug route7 (2%)1 (0.2%)10.87 (1.34 to 88.04)NA
 Kobus, 2012 [32]Long-acting opioids400 (88%)1637 (34%)2.60 (2.47, 2.74)NA
Substance use
 Chang, 2018b [4]Opioid disorders530 (19%)1243 (1%)28.95 (26.34, 31.82)NA
 Campbell, 2015 [34]Illicit drug use past 12 months71 (17%)67 (10%)11.03 (5.75 to 21.14)NA
 Kobus, 2012 [32]Substance use disorder141 (31%)1151 (24%)1.30 (1.13 to 1.51)NA
Clinical factors
 Campbell, 2015 [34]Back or neck problems344 (81%)484 (73%)1.10 (1.03 to 1.18)NA
Frequent/severe headaches134 (32%)170 (26%)1.22 (1.01 to 1.48)NA
Healthcare utilization
 Chang, 2018a [30]Opioids from ≥ 4 unique prescribers and pharmacies over 90 days1176 (0.78%)1948 (0.05%)15.6 (14.51 to 16.76)137 (129 to 145)
 Chang, 2018b [4]> 1 Hospitalizations    
 Concurrent 2012443 (16%)17,061 (9%)1.76 (1.62, 1.92)14 (12 to 18)
 Prospective 2013396 (14%)11,110 (6%)2.42 (2.21, 2.66)12 (10 to 14)
 Kobus, 2012 [32]Any pain clinic visits 6 months before/after index date104 (23%)530 (11%)2.09 (1.73 to 2.51)8 (6 to 12)
Filled opioid prescription 5 days after emergency department visit285 (63%)2696 (56%)1.12 (1.04 to 1.21)14 (9 to 46)
Mental health
 Kobus, 2012 [32]Posttraumatic stress disorder diagnostic code 309.8120 (4%)96 (2%)2.21 (1.38 to 3.55)NA
Prescribers
 Mean (%)Mean (%)RR (95% CI) 
 Chang, 2018a [30]Proportion of prescriptions from high-risk* prescribers122,159 (81%)973,865 (25%)3.24 (3.23 to 3.25)NA
Count (%)Count (%)RR (95% CI) 
100% of opioid prescriptions from high-risk* prescribers77,217 (51%)572,633 (15%)3.48 (3.46 to 3.50)NA
50–99% of prescriptions from high-risk* prescribers51,277 (34%)471,351 (12%)2.81 (2.79 to 2.83)NA
  1. CI confidence interval, ICD-10 international classification of diseases 10th revision, NNTH number needed to harm, RR relative risk, *high-risk prescribers were defined as those in the top 5th percentile of opioid volume