A STATED AND REVEALED DISEASE MODEL FOR ASTHMA

Purpose: This paper presents the development of a stated-revealed preference model integrating implicit economic information (from choice experiments on asthma cases) into a disease econometric step model; it contributes to the emerging trend of  experimental research on financials and medical decision making (Gunther et als, 2010;Vohs, Rodelmeier, 2009; ENDEP-biomed,2003) and recent relevant judgment analysis research (Seong and Bisantz, 2008; Mao and Benbasa, 2001) . It also aims to use findings in a sensitive analysis for calibration of the model. 

Method: Cost cognitive cues were identified with a focus group of 5 physicians (3 in teaching and non teaching hospitals, 2 in family practices- Boston,2006).  An internet survey of asthma physicians was run in parallel in the Kaiser Northern Medical group. The revealed preference model was designed using an extracted data set (from the 2003 NAMCS physician survey) of  223 patients diagnosed with asthma (ICD9 493, including intrinsic, extrinsic and cases of acute exacerbation). The cognitive economic information is matched to the CDC categorization of types of plans and payments coded in the survey.    

Result: Cost cues relate to economics of prescribing (combo and nebulizers, some branded drugs and polypharmacy) , physicians’economics (threat of salary cuts, reduced clinical time and more transactions) , time pressures and increasing cost of group decisions.  Relevant clinical cases for calibration were identified when physicians start thinking long term and chronic care, mainly for patients in non acute cases already stabilized. Cost aware physicians were 0.84 less likely to prescribe for asthma cases in the NAMCS category of patients on Medicare plan only, than for other Medicare  patients. Recent runs show influence of ebilling. Cost cues help to interpret unknown attributes, mainly from physicians and practices’ characteristics. They also demonstrate how decision making process moves to more group decisions (financial counselors and allied professionals).

Conclusion: A stated and revealed model can better predict treatment choices and prescribing patterns. It can integrate, over a series of years, conjoint type of micro data to control how financials influence clinical choices in period of economic transition (for very depressed sites). This method and system can help to anticipate some critical medical errors or provide adjustments in asthma cases (e.g. with COPD complications,Camargo, 2003).