Validation of the Supportive Care Needs Survey—short form 34 with a simplified response format in men with prostate cancer Penelope Schofield 1,2 *, Karla Gough 1 , Kerryann Lotfi‐Jam 1,2 and Sanchia Aranda 1,2 1 Department of Nursing and Supportive Care Research, Peter MacCallum Cancer Centre, East Melbourne, Vic., Australia 2 Melbourne School of Health Sciences, University of Melbourne, Parkville, Vic., Australia * Correspondence to: Department of Nursing and Supportive Care Research, Peter MacCallum Cancer Centre, Locked Bag 1, A’Beckett Street, Vic. 8006, Australia. E‐mail: penelope.schofield@ petermac.org Received: 20 January 2011 Revised: 24 May 2011 Accepted: 27 May 2011 Abstract Objective: The Supportive Care Needs Survey—short form (SCNS‐SF34) is a commonly used instrument that assesses levels of unmet needs in cancer patients across five domains. Partici- pants in a number of recent studies have experienced difficulties with the five‐point, two‐level response scale. This study aimed to validate the SCNS‐SF34 with a simplified four‐point response format in a large sample of Australian men with prostate cancer. Methods: Three hundred thirty‐two men with prostate cancer commencing external beam radiotherapy completed the SCNS‐SF34 with revised response format, along with measures of psychological distress (Hospital Anxiety and Depression Scale) and quality of life (Expanded Prostate Cancer Index Composite—short form). Exploratory factor and parallel analyses were undertaken to examine the structure of the revised instrument. Reliability analysis was per- formed, and convergent and divergent validity were examined using a priori predictions. Results: As with the original scale, a five‐factor solution was indicated. Four of the five factors were identical to those reported in the original SNCS‐SF34 validation study. Internal consis- tency was excellent, exceeding 0.8 for all five domains. Consistent with expectations, correlations between SCNS domains and Hospital Anxiety and Depression Scale subscales indicated moderate to large convergent relations (ranging from 0.31–0.67). Conclusions: The SCNS‐SF34 with revised response format maintained the same factor struc- ture as the original, with five domains. This analysis demonstrates that the simplified response scale does not adversely affect the psychometric properties of the instrument. The SCNS‐SF34 with revised response format is reliable, valid and more acceptable to cancer patients than its predecessor. Copyright © 2011 John Wiley & Sons, Ltd. Keywords: cancer; oncology; supportive care; needs assessment; instrument validation; prostate cancer Introduction The accurate assessment of perceived needs permits the tailoring of health service provision to meet each patient’s psychosocial and supportive care needs [1]. Increasingly, needs assessments are used to develop and evaluate the impact of psychosocial and support- ive care interventions. This will generate the evi- dence required to implement patient‐centred care. Instruments not only need to have demonstrated reli- ability and validity but also need to be acceptable to use by the target audience [2]. In our research using the original Supportive Care Needs Survey (SCNS) long [1] and short forms [3] and the Needs Assessment for Advanced Cancer Patients (NA‐ACP) [4], patients experienced the re- sponse scale as confusing and difficult to complete [5–8]. All three instruments use a five‐point, two‐ level response scale with two broad categories of need: ‘No need’ and ‘Need’. The ‘No need’ category has two subcategories: ‘No need category has three subcategories: ‘Low need’, ‘Moderate need’ and ‘High Need’. Patients report specific difficulties with the ‘No need’ subcategories, finding it hard to differ- entiate between an absence of need and needs that have already been satisfied [8]. Furthermore, multi- ple answers for items are sometimes recorded, where patients report ‘No need’ but then also select low, moderate or high need, making scoring and inter- preting difficult. Removal of the ‘No need: satisfied’ subcategory provides a simplified ordinal response scale con- sistent with that used in more recently published needs assessment instruments [9,10]. Use of a sim- plified four‐point response scale with the NA‐ACP had minimal impact on the structure and composi- tion of underlying domains [8]. Given the popularity of the SCNS‐SF34, the current study sought to Copyright © 2011 John Wiley & Sons, Ltd. Psycho-Oncology Psycho-Oncology (2011) Published online in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/pon.2016