The data acquisition system of the country's breast screening service is being implicated in the alleged delayed reporting of cancer examination results to the women concerned.
The controversy may have been an artefact of the BreastScreen units’ data acquisition system, which put misleading dates on records. When entering a date on which a woman would be notified of the test result, some of the systems used did not allow a date before the current date to be entered. So if a woman had already been sent a letter at the time the data was entered, the date had to be recorded as the current date, says Brian Cox, head of the independent monitoring group, whose examination of breast-screening procedures turned up the alleged delays.
This meant the monitoring group’s figures were not necessarily accurate. “We were unable to monitor [this aspect of the unit’s operation] properly, and we are unable to determine if in a particular case, there was a data entry problem or a genuine delay.”
Cox understands that the problem in the errant systems has now been corrected, “but there are all sorts of rough edges” still affecting the system, he says – declining to elaborate.
The Ministry of Health's national screening programme manager, Julia Peters, says there were "a couple of issues of that kind [data entry limitations] last year, but those have been resolved.
"It is a new programme, new in every respect, not just the data processing side, with four brand new providers and two others who had run pilot projects. With any new programme you expect some establishment problems."
All the [IT] systems were built in accordance with the data management manual, which specifies a minimum set of data to be acquired," she says.
The three systems were supplied by Plato CIS of Auckland, Orion Systems of Auckland and Christchurch-based Software Innovations, which supplied the majority of the units. Information manager for the project, Alan Radford, confirms that there were "some data issues", but echoes Peters in saying they have been "largely resolved".
Does this mean there are outstanding issues? "Not significant ones," he says.
Radford declines to identify which of the three systems presented problems.