Table 3.
Current and lifetime patterns of cannabis use
| Characteristic | Prescribed only | Prescribed and illicit | Illicit only | Total | Comparisons estimate (95% CI)⌠ |
|---|---|---|---|---|---|
| Cannabis use before medical use, categorical non-binary, n (%) | |||||
| Never used cannabis before medical use | 126/186 (68%)a | 63/320 (20%)bc | 195/895 (22%)bc | 384/1401 (27%) | PO – P + I: 8.6 (4.5, 16.4) |
| PO – IO: 7.5 (4.4, 13.0) | |||||
| Used ‘recreationally’ but had ≥ 1 year break before medical use | 52/186 (28%)ab | 123/320 (38%)abc | 382/895 (43%)bc | 557/1401 (40%) | PO – IO: 0.5 (0.3, 0.9) |
| PO – P + I: 0.1 (0.0, 0.2) | |||||
| Was using recreationally when started using medically | 8/186 (4.3%)a | 134/320 (42%)bc | 318/895 (36%)bc | 460/1401 (33%) | PO – IO: 0.1 (0.0, 0.3) |
| Age first tried cannabis for any reason, numeric, (N = 1405) | |||||
| Mean (SD) | 33.2 (21.5)a | 20.1 (10.7)bc | 21.7 (13.8)bc | 22.9 (15.0) | PO – P + I: 13.2 (10.0, 16.3) |
| Median (IQR) | 21 (16, 54) | 16 (15, 20) | 17 (15, 20) | 17 (15, 21) | PO – IO: 11.5 (8.7, 14.3) |
| Age first regular cannabis use any reason, numeric, (N = 1405) | |||||
| Mean (SD) | 35.4 (23.7)a | 27.0 (15.1)bc | 28.4 (17.3)bc | 29.0 (18.0) | PO – P + I: 8.4 (4.5, 12.4) |
| Median (IQR) | 33 (18, 56) | 22 (18, 33) | 23 (18, 38) | 24 (18, 39) | PO – IO: 11.5 (3.5, 10.4) |
| Age first regular cannabis use for medical reason, numeric, (N = 1402) | |||||
| Mean (SD) | 41.3 (22.6)a | 36.1 (14.7)bc | 35.5 (18.2)bc | 36.4 (18.2) | PO – P + I: 5.2 (1.2, 9.2) |
| Median (IQR) | 46 (28, 59) | 35 (26, 45) | 35 (23, 49) | 35 (24, 50) | PO – IO: 5.8 (2.3, 9.4) |
| Never used cannabis regularly for any reason, binary categorical, n (%) | 32/187 (17%)a | 13/322 (4%)bc | 63/896 (7%)bc | 108/1405 (8%) | PO – P + I: 4.9 (2.2, 11.0) |
| PO – IO: 2.7 (1.6, 4.7) | |||||
| Number of days in previous 28 used cannabis for any reason, numeric, (N = 1403) | |||||
| Mean (SD) | 21.8 (10.0)a | 24.9 (6.7)bc | 20.2 (10.2)bc | 21.5 (9.7) | PO – P + I: -3.1 (-5.2, -1.0) |
| Median (IQR) | 28 (15, 28) | 28 (12, 28) | 28 (12, 28) | 28 (15, 28) | P + I – IO: 4.7 (3.2, 6.2) |
| Number of days in previous 28 used cannabis for medical reasons, numeric, (N = 1402) | |||||
| Mean (SD) | 21.3 (10.2)a | 23.8 (8.0)b | 18.7 (10.8)c | 20.2 (10.4) | PO – P + I: -2.5 (-4.7, -0.2) |
| PO – IO: 2.7 (0.7, 4.6) | |||||
| Median (IQR) | 28 (14, 28) | 28 (24, 28) | 25 (7, 28) | 28 (12, 28) | P + I – IO: 5.1 (3.6, 6.7) |
| Estimated proportion of cannabis use for medical reasons, numeric, Mean (SD), (N = 1392) | 96.1 (14.2)a | 91.9 (14.8)b | 84.6 (21.4)c | 87.8 (19.7) | PO – P + I: 4.2 (0.0, 8.5) |
| PO – IO: 11.5 (7.8, 15.2) | |||||
| P + I – IO: 7.2 (4.2, 10.2) | |||||
| 100% of cannabis use for medical reasons, binary categorical, n (%) | 146/186 (79%)a | 177/322 (55%)b | 378/884 (43%)c | 701/1392 (50%) | PO – P + I: 3.0 (1.8, 4.9) |
| PO – IO: 4.9 (3.1, 7.6) | |||||
| P + I – IO: 1.6 (1.2, 2.2) | |||||
| Weekly cost of medical cannabis, numeric, (N = 1395) | |||||
| Mean (SD) | $79.2 ($61.6)a | $114.0 ($77.4)b | $58.6 ($65.4)c | $74.0 ($71.6) | PO – P + I: -35.5 (-49.7, -19.6) |
| PO – IO: 20.6 (7.4, 33.7) | |||||
| Median (IQR) | $60 ($35, $100) | $100 ($50, $155) | $40 ($7, $100) | $40 ($7, $100) | P + I – IO: 55.2 (44.6, 65.8) |
| Weekly cost of medical cannabis with respondents who did not pay excluded, numeric, (N = 1405) | |||||
| Mean (SD) | $81.4 ($61.0)ac | $118.0 ($76.0)ab | $73.2 ($65.4)ac | $85.9 ($70.2) | PO – P + I: -36.5 (-51.7, -21.3) |
| Median (IQR) | $65 ($40, $100) | $100 ($50, $170) | $100 ($50, $155) | $50 ($25, $100) | P + I – IO: 44.7 (33.7, 55.7) |
Superscript letters a–c indicate statistically significant between-group differences: Pairs of columns that do not share letters are significantly different from each other (p < 0.05). Missing values excluded from denominator when calculating percentages. Some percentages may not total 100% due to rounding. Letters a–c indicate statistically significant between-group differences. Columns that share letters are NOT significantly different. Median (IQR) reported for count variables only. IQR = Interquartile range. ⌠ For each outcome, pairwise comparisons were made for all three groups: PO – P + I, PO-IO, and P + I – IO. Estimates are unstandardised effects size for numeric variables and odds ratios for categorical or ordinal variables. Only significant comparisons reported