Improving patient risk profile assessment

Wednesday, February 17, 2010

According to Job van Palen, health care providers should take advantage of new research methodologies such as computerized adaptive testing. Using these methods, specialists would be better able to assess individual patients' risk profiles. 'Personalized management' would also be facilitated. Job van der Palen spoke on these topics on 11 February 2010 during his inauguration as Professor of the endowed chair in Evaluation and Assessment in Health Research at the University of Twente. He is a clinical epidemiologist at the Twente Medical Spectrum (MST).

The establishment of the new chair is a real win-win situation for MST and the University of Twente, according to Van der Palen. "The department of Research Methodology, Measurement Methods and Data Analysis (OMD) at the University of Twente has expended considerable effort applying advanced statistical methods and techniques in the field of teaching. These are techniques that can also be applied in the field of health care. In turn, hospitals need better methods to better predict how individual patients will fare during their illnesses. This has now been made possible thanks to a number of research and measurement methods that OMD has developed further - such as Item Response Theory (IRT) en Computerized Adaptive Testing (CAT)."

Consider, for example, patients who run the risk of being hospitalized within a year because of the pulmonary disorder COPD, or whose condition may be fatal. Patients who visit the pulmonary department are first given a respiratory function test. The results of this test are fed directly into the hospital's computer systems. The patient then goes to the waiting room to fill in a questionnaire. It's not a long, paper questionnaire. Computerized Adaptive Testing makes it possible to glean relevant information about the patient's condition using as few questions as possible. These results are also fed directly into the hospital's computer systems. It is only then that the patient sees the pulmonologist. The doctor consults the patient's electronic medical record and can see at a glance what the chance is that the patient will be hospitalized within a year or whether the patient's condition could potentially be fatal. Furthermore, specific factors that contribute to the patient's prognosis can also be consulted immediately. The pulmonologist can then adjust therapy accordingly. The doctor also does not have to ask the same questions over and over in the limited time available; the pulmonologist can zero in on relevant points based on the patient's responses to the questionnaire.

These research and measurement techniques may also be used to develop more accurate risk profiles for other patients, such as those with a heart condition. Some patients may be candidates for defibrillator implantation. If they suffer an repeated infarction, their chances of survival will increase thanks to the electrical shock given by the implant. Many people currently have a defibrillator implant, even though only a small minority will ever suffer a repeated infarction. Improved risk assessment will make it easier to determine which patients are the most likely candidates for an implant.

Or take patients who undergo open heart surgery. Twenty to twenty-five percent develops delirium, further degrading their condition and extending their stay in the hospital. Using reliable predictions about who will develop delirium, interventions can be considered for this risk group consisting of medication or psychological resistance training, for example.

"Improved individual care and personalized management - these are the goals of this new chair for MST and the University of Twente", according to Van der Palen.

The inaugural speech "Beter weten door beter meten" ('Knowing more through better measurements') is available in electronic format on request.

 

Palen