The implementation of guidelines and computerised forms improves the completeness of cancer pathology reporting. The CROPS project

Abstract

The aim of this study was to determine whether reporting guidelines and computerised form-based reports improve the completeness of histopathological cancer data available for patient management and population cancer registration and to evaluate the acceptability of the intervention. The study was a randomised controlled trial with a split unit design and stratified cluster randomisation. All 16 hospital pathology laboratories in Wales were randomly allocated to report either breast or colorectal resection specimens by computerised form or conventional free text. 1044 reports were analysed in the study arm, 998 in the control arm. Use of pre-defined forms led to a 28.4% (95% confidence interval (CI): 15.7–41.2%) increase in complete reporting of a minimum dataset required for cancer registration and a 24.5% (95% CI: 11.0–38.0%) increase in complete reporting of minimum data required for patient management. Form-based reporting was acceptable to pathologists and preferred by clinicians. In conclusion, guidelines and computerised forms significantly improve the quality of histopathology reporting.

Introduction

Timely, complete and accurate pathology reports are fundamental to the provision of quality cancer services. Inconsistent pathology reporting, such that essential data items of therapeutic or prognostic relevance are sometimes missing, can lead to inconvenience for the patient and the clinician, delays in treatment, and inadequate or inappropriate postoperative therapy. Consequently, pathology reporting contributes to differences in standards of treatment between clinical teams. Previous audits carried out in Wales on breast and colorectal cancer histopathology reporting showed significant inter- and intra-hospital variation in the data items mentioned in reports, indicating scope for standardisation. For example, in 1991 only 55% of pathology reports on invasive breast cancer specimens contained information on histological grade [1] when there is strong evidence that grade predicts survival and response to chemotherapy 2, 3, 4, 5. In 1993, only 51.5% of rectal cancer reports contained a statement on the completeness of excision at the circumferential margin [6] when this predicts local recurrence and may indicate the need for post-operative radiotherapy [7]. There is evidence to show that the use of standardised pre-defined forms (as opposed to the use of free text) improves the quality and completeness of pathology reports 8, 9. However, these studies have only been carried out in specialist teaching hospitals, and the generalisability of the findings to district hospital laboratories has not been investigated.

Histopathology data also form an important element of cancer registration and accurate, complete recording of pathology information contributes to the overall quality of a registry by facilitating high levels of ascertainment and satisfying national requirements for data completeness, including pathological staging. Currently data entry in most registries depends on the interpretation of free text pathology reports by trained coders. The reliability of coding is affected by the fact that pathology reports are actually intended for communication with clinicians making therapeutic decisions rather than clinical coders who work to objective rules. Free text pathology reports therefore raise data quality issues from a Registry perspective because they are not designed for cancer registration purposes. The potential for collecting histopathology data electronically and its direct transmission to the Cancer Registry in coded format, obviating the need for any interpretative stage, has been recognised for some time [10], but there has been little attempt at its implementation on a regional or national scale.

We report here the findings of the Cancer Registration through On-line Pathology Systems (CROPS) project in Wales, UK. This project was funded in 1996 when the (then) Welsh Office of the UK Government was implementing an all Wales pathology computer system. This provided an opportunity to investigate the use of standardised computerised forms containing a set of specified data items, supplemented by written and on-screen guidelines, for reporting cancer specimens in all 16 hospitals in the Principality, and to consider the possibility of using electronic transfer of this standardised pathology data to the Cancer Registry.