Skip to main content

Table 4 Causes underlying negative and positive false cases in prostate cancer (with regard to the highest predictive algorithm)

From: Is hospital discharge administrative data an appropriate source of information for cancer registries purposes? Some insights from four Spanish registries

  Granada
Dataset
n (%)
Basque Country
Dataset
n (%)
Murcia
Dataset
n (%)
Zaragoza
Dataset
n (%)
False negative cases     
• Cases from private hospital in 2000 27 (12.8) 372 (37.4) 0 (0) 25 (6.8)
• Cases admitted in 2000 but discharged in 2001 3 (1.4) 98 (9.8) 42 (11.1) 43 (11.7)
• Cases coded in dx6 position or more - 7 (0.7) 3 (0.8) -
• False negative diagnoses 7 (3.3) 46 (4.6) 3 (0.8) 2 (0.5)
• Ambulatory care (instead of in-hospital care) 147 (69.7) 343 (34.4) 262(69.3) 195(53.1)
• Loss using the algorithm 25 (11.8) 125 (12.6) 52 (13.7) 99(26.9)
• Other causes 2 (0.9) 4 (0.4) 16 (4.2) 3 (0.8)
  211 (100) 995(100) 378(100) 367(100)
False positive cases     
• Cases registered in 2001 - 2 (2.1) 1 (3.8) -
• Prevalent cases 9 (69.2) 78 (82.1) 14 (66.7) 36 (81.8)
• False positive diagnoses 4 (30.7) - 4 (15.4) 6 (13.6)
• Missing in the registry - 4 (4.2) 0 (0) -
• Other residence than that covered by the registry - 7 (7.4) 1 (3.8) 2 (4.5)
• Other causes - 4 (4.2) 6 (23.1) -
  13(100) 95(100) 26(100) 44(100)
\