Title: Risser Patient Satisfaction Scale: A Validation study in Greek Cancer Patients

Author's response to reviews Title: Risser Patient Satisfaction Scale: A Validation study in Greek Cancer Patients Authors: Andreas I Charalambous (andreas.charalambous@cut.ac.cy) Theodoula Adamakidou (adamakidou@yahoo.gr) Version: 2 Date: 29 October 2012 Author's response to reviews: see over

29/10/2012 Editor-in-Chief Dear Editor Enclosed for your consideration is the revised version of the original article, entitled "Risser Patient Satisfaction Scale: Cultural adaptation and validation study in Greece". The authors have read the points raised by the reviewers and mutually decided to fully addressed these. In order to succeed this, the authors had to undertake extensive revision of the manuscript with regards to the points raised by the reviewers. The changes are presented as point-by-point response to the reviewers suggestions. Editorial comments Please include your title page at the beginning of your manuscript, rather than uploading it as an additional file DONE Reviewer Flora Tzelepis Changes made as per her comments: 1. Abstract, Methods: Include a in first sentence so reads This is a test retest DONE 2. Abstract, Conclusion: Delete the apostrophe so reads Greek version offers DONE 3. Background, paragraph 1: Include a comma after characteristics so reads patient characteristics, attitudes DONE 4. Background, paragraph 1: Delete it so reads between what the patient expects and what is offered by the nursing care.- DONE 5. Background, paragraph 2: Change even thought this.. to even though this.. DONE 6. Methods, paragraph 1: Were 18 year olds eligible for the study? If so, change >18 years to 18 years or older. CHANGED TO 18 years or older. 7. Methods, paragraph 1: Change that they to who so reads Eligible participants were adult cancer patients who were receiving DONE 8. Methods, paragraph 3: Add 2010 after November so reads November 2010 INSERTED 9. Methods, paragraph 3: Delete was so reads the final sample consisted of DELETED 10. The paragraph was changed to read as follows: Out of the 326 patients that were identified in the selection process, 28 patients did not meet the inclusion criteria. The twenty-eight patients were excluded for the following reasons: 3 (0.92%) refused to participate in the study, 5 (1.55%) had brain metastasis with impaired cognitive ability, 10 (3.06%) were admitted in less than 2 days in the hospital, 4 1

(1.22%) had communication problems, and 6 (1.84%) were in the terminal stage of the illness. Therefore, the final sample consisted of 298 participants who all answered the questionnaire (response rate 100%).. 11. Methods, paragraph 4: Change detail to detailed so reads receiving detailed oral and written CHANGED TO DETAILED 12. Methods, paragraph 5: Suggest replacing the last statement with something like Of the 298 patients provided with a re-test survey, (insert number) (insert %) completed this second questionnaire. CHANGED TO Of the 298 patients provided with a re-test survey, 253 (85%) completed this second questionnaire 13. Translation and Cultural Adaptation, paragraph 1: Suggest changing review to reviewed and is to was so statement reads They reviewed the questionnaire and were asked specific questions by the researchers as to whether the translation was CHANGED 14. Translation and Cultural Adaptation, paragraph 1: Does the statement Finally, the pretest and reliability testing (test-retest) followed. relate to the 15 patients or does this sentence relate to the main study described in the paper. THIS REFERS TO THE MAIN STUDY. 15. Data analysis, final paragraph: Include a. after more reliable. ADDED 16. Results, paragraph 1: Should 158 (53%) of the patients had 2-10 days after admission read 158 (53%) of the patients were hospitalised for 2-10 days. CHANGED TO 158 (53%) of the patients were hospitalised for 2-10 days 17. Results, paragraph 3: Include a. after second highest. ADDED 18. Results and Discussion, final paragraph: The authors indicate that the PSS scale has already been translated into Greek. Could the authors have used this previously translated scale rather than have the scale translated again? NO IT WAS NOT POSSIBLE TO USE THE PREVIOUS GREEK VERSION AND THE RATIONALE IS ANALYSED IN THE TEXT AS FOLLOWS The PSS Scale has been used world widely across different cultures. It has been used in different non-english speaking countries and translated in several languages. The scale has been previously translated in Greek, however the adaptation and validation was tested in Cypriot population and not Greek. Despite the many apparent commonalities between the population of the two countries there are also not so apparent discrepancies that call for suspicion when validated instruments in one country are about to be used in the other and vice versa. Therefore it was considered necessary by the researchers, to adapt this scale explicitly for the Greek population and the results verified their decision as differences were found between the two available Greek versions. 19. Limitations paragraph 1: The authors indicated that the research took place in one of the three oncology hospitals in Athens. Was the research undertaken in the largest hospital? YES IT IS THE LARGEST HOSPITAL IN THE ATHENS REGION 20. Table 1: Instead of, could use. in the % for example 44.6 instead of 44,6 CHANGED 21. Table 3: Gronbach should read Cronbach and could change the, in this column to CHANGED 22. Table 4: Could change the, in the Coefficient Alpha column to CHANGED 2

23. The section Results and discussion was divided into two separate parts. Reviewer A Philalithis Changes made as per comments: 1. This point raised by the reviewer is answered in detail at the Results and Discussion section and explicitly by the following paragraph The PSS Scale has been used world widely across different cultures. It has been used in different non-english speaking countries and translated in several languages. The scale has been previously translated in Greek, however the adaptation and validation was tested in Cypriot population and not Greek. Despite the many apparent commonalities between the population of the two countries there are also not so apparent discrepancies that call for suspicion when validated instruments in one country are about to be used in the other and vice versa. Therefore it was considered necessary by the researchers, to adapt this scale explicitly for the Greek population and the results verified their decision as differences were found between the two available Greek versions. The translation and cultural adaptation in the two populations indeed raised issues of understanding and comprehensiveness between the available two translations. For example if we consider the question The nurse is understanding in listening to a patient s problems, this was translated differently in the two populations in order to achieve the same meaning. Explicitly in the one version the question was translated as «Η νοζηλεύηρια επιδεικνύει καηανόηζη ζηα προβλήμαηα ηοσ αζθενή» and on the other version was translated as «Η νοζηλεύηρια καηανοεί ηα προβλήμαηα ποσ ανηιμεηωπίζει ο αζθενής». 2. The title of the paper has been revised to Risser Patient Satisfaction Scale: A Validation study in Greek Cancer Patients. The authors agree with the reviewer that the title over-emphasized this aspect of the study, which was not the authors intention. It is clarified, that the adaptation and validation refers to the Greeklanguage version of the questionnaire. A sentence was revised to clarify this point: This study will allow cross-cultural adaptation and validation of the Greek-language version of the questionnaire in the Greek population with its distinctive cultural influences. 3. The first paragraph of the Translation and Cultural Adaptation section was revised accordingly in order to address the points raised by the reviewer: Part of this study was the translation and cultural adaptation of the Risser questionnaire in the Greek language. Although, a Greek version was available, the translation and adaptation occurred in a different population (Greek-Cypriot). Although Greek- Greek vernacular differs from Cypriot-Greek vernacular several changes were necessary in order to adjust the questionnaire to the Greek (Athenian) patients. For example the question The nurse is understanding in listening to a patient s problems, this was translated differently in the two populations in order to achieve the same meaning. Explicitly in the Greek-Cypriot version the question was translated as «Η νοζηλεύηρια επιδεικνύει καηανόηζη ζηα προβλήμαηα ηοσ αζθενή» and in the Greek-Greek version was translated as «Η νοζηλεύηρια 3

καηανοεί ηα προβλήμαηα ποσ ανηιμεηωπίζει ο αζθενής». The adaptation and translation process was the same in the two countries with one exception. In the Greek study (paper under review), 3 independent bilingual nurses were used for the English-Greek translation and another 3 for the back translation. However, this is indicated by the Minimal Translation Criteria, adopted in this study. The different numbers in the independent bilingual translators used, does not change the essence of the process. In order to make the translation process more comprehensive to the reader the authors have also revised the second paragraph of the Translation and Cultural adaptation section to read as follows In order to produce an adapted questionnaire of the highest semantic equivalence it is important to follow internationally recommended criteria suggested by the relevant literature [29, 30]. Therefore, the adaptation process was based on the Minimal Translation Criteria [31] that included translation and back translation of the original questionnaire. Three independent bilingual nurses with previous experience in translating questionnaires produced the English to Greek translation. Subsequently, the questionnaire was back-translated in Greek by three independent bilingual nurses. The produced English versions of the questionnaire were compared with the original one, and this process identified some problematic questions. These were addressed by revising the questions based on the translators mutually agreed suggestions. 4. In order to address this point raised by the reviewer a new table (numbered as table 1) was added to the paper (at the section Questionnaire Description ) that presents the items of the questionnaire (all 25 ones) along with their computed means and Standard Deviations. 5. As with the Cypriot study, the Greek study used the same version of the Risser questionnaire. This is the version that differs very little compared to the original one, and it is the one developed by Hinsaw and Atwood. The only differences between the two, is the 7 th item of the technical-professional subscale where the phrase over the telephone was deleted. The acceptability and appropriateness of this questionnaire for tertiary care has been acknowledged and tested in previous studies such as the Hinshaw and Atwood studies (5 in total) or the Rafii et al. study that primarily used inpatients as their sample. Of course there is also the Cypriot study that also used inpatients in cancer care settings which also in favour of the applicability and appropriateness of this questionnaire in tertiary settings. Furthermore, the first paragraph of the questionnaire description section was revised to read as follows A modified version of the Risser [8] Patient Satisfaction Scale (PSS) was used to elicit the research data. The version implemented here is the one produced by Hinshaw and Atwood [50] which compared to the original Risser scale differs at the 7 th item of the technicalprofessional subscale where the phrase over the telephone was deleted. This version was psychometrically tested in five studies with a total of 600 patients, primarily medical-surgical inpatients. The results showed stable internal consistency estimates in the different studies with the average coefficients alpha values reported being 0.79, 0.78 and 0.88 for the three subscales respectively. The PSS was designed to evaluate patients attitudes towards nurses and nursing, 4

and originally contained three subscales with a total number of 25 items (Table 1) defined as follows [ ] 6. The reviewer comments on the following sentence The view expressed by Apolone and Mosconi [24] that whether we really need new questionnaires has been the guide to adapt, test and validate a questionnaire for patient satisfaction in Greek. There has been a misunderstanding here, the authors intentions was not to suggest that they have adopt the same process as this was described in Apolone and Mosconi editorial. There citation was to support that view that indeed more satisfaction questionnaires are needed in the process of evaluating patients outcomes. In order to clarify the true intention, the authors have revised this sentence as follows This perspective on the necessity of satisfaction scales in the enhancement of patients outcomes evaluation has been stressed by Apolone and Mosconi [24] and this study comes as a response to the need to adapt, test and validate questionnaires for patient satisfaction in Greek.. 7. A. In order to clarify this point the following sentence was added: The sampling frame that was employed included consecutive patients admitted in the whole hospital. B. The questions and clarifications requested by the reviewer required major revision to the section, particularly the first three paragraphs that were revised as follows: The study was conducted in a large Anticancer Hospital in Athens. Potential eligible participants were identified prior to running the random number selection program based on a set of pre-determined inclusion and exclusion criteria. The inclusion criteria included adult cancer patients 18 years or older who were receiving care at the hospital for at least 48 hours. The potential participants needed to be able to speak and understand Greek and they had accepted to provide a written informed consent. No restrictions were imposed in relation to the type of cancer. Finally prospective eligible patients should have a score of >50 on the Karnofsky Performance Scale Index [27] and a mean of >50 on the Attentional Function Index (AFI) which was used to measure perceived cognitive function [28]. Patient's performance was assessed by a research assistant and the AFI was assessed by the patients themselves prior to running the random number selection program. The results of these assessments varied for both scales. For the Karnofsky Scale the results ranged between 60 and 80 with the level of 60 indicating that the patient Requires occasional assistance, but is able to care for most personal needs and the level of 80 indicating that the patients has Normal activity with effort; some signs or symptoms of disease [27]. The AFI produced scores ranging from 55-82. Patients who score 50 to 75 function moderately well and patients who score >75 function well [29]. The exclusion criteria included patients at a terminal phase of the illness (receiving palliative care), impaired cognitive ability, patients diagnosed within the last 6 months and quarantined patients (transplantation, infections). A random number selection program was used for the selection of the patients over a period of 3 months (extending from November 2010 to January 2011). The sampling frame that was employed included consecutive patients admitted in the whole hospital. Out of the 326 patients that were identified as potential 5

participants during the selection process, 28 patients were further excluded for various reasons: 3 (0.92%) refused to participate in the study, 5 (1.55%) had brain metastasis with impaired cognitive ability (deteriorated after admission), 10 (3.06%) received an early discharge or transfer to another hospital (cared for less than 2 days in the hospital), 4 (1.22%) had communication problems and 6 (1.84%) were in the terminal stage of the illness C. Patient's performance was assessed by a research assistant and the AFI (Attentional Function Index) was assessed by the patients themselves (as this is a self-reported scale). D. The follow up (re-test) was confidential and anonymous as the two questionnaires (test and retest) from the same respondent were compared based on unique pairing codes. The identity of the respondent could not be disclosed. A clarification already is provided in the text Participants were provided with stamped envelopes and the anonymous questionnaire, and were asked to post it once they have completed 4 weeks after their discharged. Call and text reminders were made to all the participants in order to assure a high response rate. In the brackets the phrase ((with a pairing code)) was added following the word questionnaire in order to indicate that the questionnaires are paired with the ones administered during the test phase. E. The last sentence of this section was rephrased to answer this point. The sentence now reads as Of the 298 patients provided with a re-test survey, 253 (85%) completed this second questionnaire.. 8. The section results and discussion was divided to two separate parts namely results and discussion as per the reviewer s comments. The second paragraph of the discussion section was revised in order to accommodate some comments removed from the results section as recommended by the reviewer: The study s findings revealed that the Greek version of the Risser Patient Satisfaction Scale is a valid and reliable tool that is appropriate to elicit data on cancer patients satisfaction with the received nursing care. The total Alpha coefficient as well as the individual alphas of the subscales are >70 that signify a very good reliability of the scale. The reported alphas are well above the international recommended minimum criterion and comparable to the previous validation studies [26, 58, 59, 60]. The scale also demonstrated significant criterion validity that coincides with those found in the Cypriot study [26]. 9. As suggested by the reviewer the following paragraph of the limitations section was moved to the discussion section. The paragraph was also slightly modified and literature references were used as to make the argument stronger As patients had a number of different nurses caring for them, they had problems answering questions that referred to all nurses. In addition, cancer nursing occurs within a multidisciplinary or interdisciplinary context and patients had difficulties to isolate the nursing care from the whole health care experience. Therefore, this aspect raises concerns if indeed patients perceived being satisfied or dissatisfied based on the nursing interventions and interactions. The lack of sensitivity and the difficulty of the patients to distinguish nursing care from their overall experience with health care posses as a threat to the validity of these 6

measurements [61, 62]. However, this is not a new problem but rather an aspect that satisfaction scales have failed to address over time [62]. 10. The limitation section has been re-writen in order to address the concern raised by the reviewer. The revised section reads as follows: The Greek version of PSS demonstrated psychometric properties comparable to those reported for the original version in other European countries; however, a number of limitations need to be acknowledged for this study. The research study was undertaken in single anticancer hospital in Athens. However, taking into consideration that similar conditions exist in the other three anticancer hospitals, it is possible that the findings can be generalizable. Moreover, the fact that this was a validation study and not a study aiming to actually measure patients satisfaction, generalizability is not really an issue nor it posses a threat to the findings. Another limitation was that patients may have felt constrained in their responses if they perceived that this information may be provided to their healthcare providers [64]. For the same reason, patients might have favoured the positive responses to the questions. According to the study protocol, the questionnaire was distributed only to patients based on pre-determined inclusion and exclusion criteria, therefore, patients with a different background might have responded differently to the questionnaire. Minor Essential Revisions 1. in health care industry (replaced) in the health care sector 2. has been amplified (replaced) has been intensified 3. Therefore in order to (replaced) Therefore, in order to 4. what it is offered by the nursing care (replaced) what is offered 5. even thought (replaced) even though 6. Nowadays, nursing care is recognized as an area subjected to competition (replaced) Nowadays, nursing care is recognized as an area of healthcare where 7. an aspect that previously was let unexplored an aspect that previously was left unexplored 8. The PSS has served all of Rubin s criteria The PSS has satisfied all of Rubin s criteria 9. that they were receiving care at the who were receiving care at the 10. the final sample was consisted of 298 the final sample consisted of 298 11. verbiage wording 12. convenient convenience 13. (Altman, 1991; Fleiss et al, 2004; Shrout, 1998) replaced with [39,40,41] 14. Gastrointestinal cancer was the most prevalent diagnosis Gastrointestinal cancer was the most common diagnosis 15. less that 1 less than 1 16. The PSS Scale has been used world widely The PSS Scale has been used world-wide 17. translated in several languages translated into several languages 18. Mor V, Laliberle L, Wiemann Mor V, Laliberte L, Wiemann 7

Yours Sincerely Dr. Andreas Charalambous Dr. Theodoula Adamakidou 8