Introduction A 12-item scale, which relates to patients' intrinsic desire for information (IDI), was originally extracted from a larger 50-item exploratory survey questionnaire administered to 501 patients.¹ This study aimed to refine the IDI scale, by interfacing quantitative and qualitative data in a further patient sample, to develop a potentially valuable tool for predicting the specific drug information needs of individual patients. The results presented here provide a reliable qualitative description of the raw scores obtained from the IDI scale.

Method Standardised bedside interviews were conducted on 10 general medical wards at two teaching hospital trusts. Patients were sequentially recruited over an eight-week period. Data included demographic detail, the IDI items and a series of open qualitative questions, which were transcribed at the bedside. Responses to the original 12-item IDI questionnaire were scored on a five-step Likert scale, ("strongly disagree" to "strongly agree"). Quantitative and qualitative data were coded and entered on appropriate software (SPSSv10 and NUD*ISTv4 respectively). Qualitative data were subject to content analysis and then interfaced with the scale scores.

Results Two-hundred-and-ninety-nine patients were entered. The 12 items of the original IDI scale were subject to principal components analysis with oblique rotation; the structure matrix suggested two secondary factors, including a six-item construct, identified as "the extent of information desired (EID)". This reliable construct (coefficient alpha = 0.7) allowed individual patient scores and quartile ranges to be calculated which were interfaced with the qualitative data content analysis (Table 1). The coding constructs from the content analysis were independently tested for reliability (kappa range 0.66 to 0.94, p<0.001). Most patients requested basic information, such as when and how to take their prescribed drugs. However, differences emerged between the first and fourth quartile groups; the low scoring, first quartile group had more anxiety about how and if the medicine would improve their condition; particularly they wanted reassurance of benefit. This was less important to the high scoring, fourth quartile group who focused on side effects and how to deal with them. The issue of side effects was not readily apparent in the low scoring group who were characterised by notions of "trust" and "faith" which mitigated concerns about side effects.

Focal points

• The methodology employs a novel method of interfacing qualitative and quantitatively derived data in order to validate a new construct centred on a patient's intrinsic desire for drug and medical information
• The cohort size (n=299) and the reliability of the coding process is sufficient to provide a robust result for the further development of this diagnostic tool
• Patients vary in their information desires. Those with low desires tend toward more reassurance about benefit and have more faith and trust in the health care professionals involved with their treatment
• Patients with higher levels of information need can be targeted on admission and pharmaceutical care services delivered appropriate to this need
• By providing evidence for the anchor points of the construct, raw scores can be used direct at the bedside to help with information need planning