Outside the high-volume unit: survival in an 11-year cohort of colorectal cancer patients.

INTRODUCTION: The policy of redistributing surgical case volume toward designated high-volume hospitals to improve outcome in cancer is supported by an international literature on volume-outcome association. METHODS: All patients who underwent surgery for colorectal carcinoma under the care of one surgeon at a non-high-volume hospital 1995-2005 were identified. 5-year overall survival probability and 30-day operative mortality were measured. RESULTS: Two hundred and forty patients were identified. Mean annual surgeon caseload was 21.6 (SD 4.2). 5-year overall survival probability was 57.1% (95% confidence interval ±7.4%). 30-day operative mortality was 4.6%. CONCLUSION: Estimates of outcome were not different from publically available values from a high-volume unit in Ireland. These findings suggest that concentrating case volume per se may not improve outcome to the extent desired. Future improvement in colorectal cancer outcome is just as likely to derive from wider screening, better surgical training, and adequately powered clinical research, should these accompany centralisation.

Clinical observed performance evaluation: a prospective study in final year students of surgery.

We report a prospective study of clinical observed performance evaluation (COPE) for 197 medical students in the pre-qualification year of clinical education. Psychometric quality was the main endpoint. Students were assessed in groups of 5 in 40-min patient encounters, with each student the focus of evaluation for 8 min. Each student had a series of assessments in a 25-week teaching programme. Over time, several clinicians from a pool of 16 surgical consultants and registrars evaluated each student by direct observation. A structured rating form was used for assessment data. Variance component analysis (VCA), internal consistency and inter-rater agreement were used to estimate reliability. The predictive and convergent validity of COPE in relation to summative OSCE, long case, and overall final examination was estimated. Median number of COPE assessments per student was 7. Generalisability of a mean score over 7 COPE assessments was 0.66, equal to that of an 8 × 7.5 min station final OSCE. Internal consistency was 0.88-0.97 and inter-rater agreement 0.82. Significant correlations were observed with OSCE performance (R = 0.55 disattenuated) and long case (R = 0.47 disattenuated). Convergent validity was 0.81 by VCA. Overall final examination performance was linearly related to mean COPE score with standard error 3.7%. COPE permitted efficient serial assessment of a large cohort of final year students in a real world setting. Its psychometric quality compared well with conventional assessments and with other direct observation instruments as reported in the literature. Effect on learning, and translation to clinical care, are directions for future research.

Estrogen receptor-beta mRNA is associated with adverse outcome in patients with breast cancer.

BACKGROUND/AIMS: Estrogen receptor (ER) is the prototype therapy predictive marker in oncology. The ER is now known to exist in two main forms with similar overall structure: ER-alpha and ER-beta. Both forms may be expressed in breast cancer. The aim of this study was to examine breast cancer outcome in relation to expression of ER-beta. METHODS: In this investigation, we measured the expression of ER-alpha protein and ER-beta mRNA in 121 extracts of invasive breast cancer. Association of expression with clinical outcome was examined using Kaplan-Meier and Cox regression analyses. RESULTS: While ER-alpha expression was associated with good patient outcome [hazard ratio (HR) for death from breast cancer 0.37; 95% confidence interval (CI) 0.17-0.84; p = 0.017], ER-beta predicted poor outcome (HR for death from breast cancer 2.49; 95% CI 1.10-5.63; p = 0.028). CONCLUSION: Based on these findings, we conclude that ER-beta may have a different biological role from that of ER-alpha in breast cancer.