Table 5.
Factors significantly associated with the propensity of GPs to testing asymptomatic men
| Yes | Total | Univariate | Multivariate | |||
| Num | % | Num | OR [95%CI] | OR [95%CI] | ||
| GP Characteristics | ||||||
| Gender | Male | 867 | 79.7 | 1,088 | 1.00 | 1.00 |
| Female | 408 | 79.1 | 516 | 0.96 [0.74,1.25] | 1.49 [1.08,2.05] | |
| p = 0.016 | ||||||
| Age group | < 50 years | 604 | 74.3 | 813 | 1.00 | 1.00 |
| ≥ 50 years | 660 | 85.1 | 776 | 1.97 [1.53,2.54] | 1.71 [1.27,2.31] | |
| p = 0.002 | ||||||
| Full-time principal | Yes | 1,038 | 81.4 | 1,275 | 1.00 | 1.00 |
| No | 225 | 71.9 | 313 | 0.58 [0.44,0.78] | 0.57 [0.41,0.81] | |
| p = 0.006 | ||||||
| Length of Practice1 | ≤ 10 years | 235 | 70.6 | 333 | 1 | 1 |
| 11–20 years | 366 | 75.9 | 482 | 1.32 [0.96,1.80] | 1.45 [1.02,2.07] | |
| 21–30 years | 406 | 85.5 | 475 | 2.45 [1.73,3.47] | 2.39 [1.60,3.56] | |
| ≥ 30 years | 248 | 86.1 | 288 | 2.59 [1.72,3.89] | 2.51 [1.57,4.03] | |
| p < 0.001 | ||||||
| Special interest in: | ||||||
| Research | No | 1,251 | 80.2 | 1,559 | 1.00 | 1.00 |
| Yes | 24 | 53.3 | 45 | 0.28 [0.15,0.51] | 0.39 [0.20,0.77] | |
| p = 0.008 | ||||||
| Cancer detection | No | 1,166 | 78.3 | 1,489 | 1.00 | 1.00 |
| Yes | 109 | 94.8 | 115 | 5.03 [2.19,11.55] | 3.61 [1.48,8.85] | |
| p = 0.001 | ||||||
| Practice characteristics | ||||||
| Occupational Health Assessments | No | 342 | 72.6 | 471 | 1.00 | 1.00 |
| Yes | 907 | 82.6 | 1,068 | 1.79 [1.39,2.31] | 1.68 [1.27,2.22] | |
| p = 0.001 | ||||||
| Actively involved in GP teaching | No | 865 | 82.5 | 1,049 | 1.00 | 1.00 |
| Yes | 392 | 73.4 | 534 | 0.59 [0.46,0.75] | 0.61 [0.46,0.80] | |
| p = 0.002 | ||||||
| Run'Well man' clinic/appointments | No | 541 | 75.2 | 1,311 | 1.00 | 1.00 |
| Yes | 721 | 83.4 | 278 | 1.91 [1.31,2.77] | 1.59 [1.06,2.39] | |
| p = 0.064 | ||||||
| Have a policy on PSA testing | No | 865 | 75.2 | 719 | 1.00 | 1.00 |
| Yes | 392 | 83.4 | 865 | 1.67 [1.31,2.13] | 1.45 [1.11,1.90] | |
| p = 0.007 | ||||||
| Knowledge | ||||||
| PPV PSA level | < 30% | 472 | 71.6 | 659 | 1.00 | 1.00 |
| Overestimated (> 30%) | 743 | 84.5 | 879 | 2.16 [1.68,2.78] | 2.02 [1.55,2.63] | |
| p < 0.001 | ||||||
| PPV PSA&DRE | < 50% | 310 | 69.8 | 444 | 1.00 | 1.00 |
| Overestimated (> 50%) | 912 | 83.0 | 1,099 | 2.11 [1.63,2.72] | 2.06 [1.56,2.73] | |
| p < 0.001 | ||||||
| Clinical Practice | ||||||
| Refer 55 yr man at PSA level | ≤ 7 ng/ml | 803 | 82.9 | 969 | 1.00 | 1.00 |
| ≥ 7 ng/ml | 332 | 74.9 | 443 | 0.62 [0.47,0.81] | 0.60 [0.45,0.81] | |
| Lab reference range | 129 | 73.7 | 175 | 0.58 [0.40,0.84] | 0.61 [0.40,0.93] | |
| p = 0.002 | ||||||
| Frequency of PSA testing: Annually or less | Yes | 457 | 94.2 | 485 | 6.01 [4.01,9.01] | 5.54 [3.64,8.45] |
| No | 793 | 73.1 | 1,085 | 1.00 | 1.00 | |
| p < 0.001 | ||||||
| Perform other routine tests with PSA | No | 764 | 85.6 | 893 | 1.00 | 1.00 |
| Yes | 505 | 71.6 | 705 | 2.35 [1.83,3.01] | 1.97 [1.51,2.57] | |
| p < 0.001 | ||||||
| Inform patient prior to testing | Frequently | 1,193 | 81.9 | 1,457 | 1.00 | 1.00 |
| Sometimes | 69 | 68.3 | 101 | 0.48 [0.31,0.74] | 0.53 [0.31,0.89] | |
| Rarely/Never | 9 | 27.3 | 33 | 0.08 [0.04,0.18] | 0.09 [0.04,0.23] | |
| p < 0.001 | ||||||
| Had an asymptomatic man diagnosed with PCa via | ||||||
| PSA testing | No | 338 | 64.4 | 525 | 1.00 | 1.00 |
| Yes | 923 | 87.0 | 1,061 | 3.70 [2.87,4.76] | 3.31 [2.51,4.37] | |
| p < 0.001 | ||||||
| Practice changed post PSA-detected PCa2 | No | 265 | 78.6 | 337 | 1.00 | 1.00 |
| Yes | 558 | 91.5 | 610 | 2.92 [1.98,4.29] | 2.78 [1.81,4.25] | |
| p < 0.001 | ||||||
| Consider having a PSA test yourself3 | No | 48 | 37.8 | 127 | 1.00 | 1.00 |
| Yes | 807 | 85.7 | 942 | 9.80 [6.56,14.65] | 8.25 [5.26,12.93] | |
| p < 0.001 | ||||||
The multivariate model was fully adjusted for characteristics of the GP (age, gender, full-time principal, health board, whether they had qualifications in geriatric medicine, palliative care or occupational health, whether they trained in UK, whether they had a specialist interest in men's health, research, urological problems or cancer detection) and the practice (solo vs multi-doctor practice, percent of patients with private health insurance, percent of time performing Occupational Health Assessments, involvement in GP teaching, running a "well man" clinic, and having a practice policy on PSA testing).
1Length of time in practice was not adjusted for age because they were correlated.
2Among GPs who had an asymptomatic man diagnosed with prostate cancer via PSA testing (n = 1,061)
3Among male GPs (n = 1,088)
PPV = Positive Predictive Value.
PCa = Prostate Cancer
DRE = Digital Rectal Exam