Supportive Care Needs Survey Short Form-34(SCNS-SF-34) Nepali Version: Psychometric Assessment Among Cervical Cancer Patients in Nepal

Purpose Psychometrically valid and reliable supportive care need survey(SCNS) instrument explores perceived supportive care needs of cervical cancer patients in comprehensively. No Nepali validated version so we decided to test the psychometric properties of the translated Nepali version of SCNS-SF 34-N(Nepali) among cervical cancer patients. Methods 334 participants were recruited purposively from 5 cancers specic hospital. Factor structure was assessed by using Exploratory factor analysis (EFA). Structure validity, Internal consistency convergent validity, and discriminant validity of the resulting factor structure were calculated and conrmed. Results Using EFA a ve-factor structure was developed considering higher loading factor for multiple loaded items which was similar with the dimensions of the original version of the SCNS-SF34 (psychological, health system and information, physical and daily living, patient care and support, and sexuality), accounting for 65.48% of the total variance. Internal consistency was achieved at an acceptable level, with Cronbach's alpha coecients ranging from 0.789 to 0.929for all ve domains and 0.887 for the whole scale. Convergent validity was conrmed by signicant inter item correlations with all corresponding items. Independent 't' test between known subgroups on age, marital status, stage of disease and treatment modalities conrmed discriminant validity. It was conrmed that the SCNS-SF 34 N(Nepali) is a valid and reliable instrument for the assessment of the supportive care needs of cervical cancer patients in Nepal


Background
Worldwide Cervical Cancer is one of the most common cancer disease in women [1,2] and 1.4 million women are living with Cervical cancer [2].The new cases of cervical cancer are approximately 570 000 and the death cases are 311 000 from this disease and it is the fourth most common cancer in women standing after breast cancer (2·1 million cases), colorectal cancer (0·8 million) and lung cancer (0·7 million) [3]. In the South East Asia region, cervical cancer is the second most common type of cancer in women and it is the main cause of cancer related death in low and middle-income countries (LMICs) like Nepal [4]. The most frequent common cancer among women in Nepal is cervical cancer. The new cases of this disease are 2,332 and death cases are 1,367 every year. It positions as the rst common malignancy among women and it commonly occurs between 15 and 44 years of age [5].
Healthcare providers can identify the patients' desire for actual services or resources in satisfying physical and daily living, psychological, sexual, patient care as well as health system and information needs through the assessment of Supportive care needs of cancer patients. When the patient recognizes a de ciency of support or care, unmet supportive care needs always occurs according to the patient's perception [6,7].
Need assessment is a critical step to provide excellent care and achieve cancer patients' and family expectation by the health care provider. Inadequate understanding of patient and family needs may result in increased unnecessary suffering and health care costs [1].
Numerous cancer patients get the experience of compromised quality of life during the disease course it may be either due to disease itself or treatmentrelated side effects. The main objective of cancer care is to maintain or improve quality of life. Maintenance or improvement of quality of life is identi ed as the main objective of cancer care [8,9].
Cervical cancer patients encountered with many physical, psychological, social distress, spiritual suffering fatigue, irritability, memory loss, decreased energy level, recurring pain and they have been faced with worse emotional distress and quality of Life in Comparison with other gynaecological cancers. Supportive care helps the patient and family to handle the problems of the illness in all phase of the disease in line with holistic health care management [10,11].
Psychometrically strong instruments are essential for the assessment of supportive care needs of cancer patient holistically in the area of clinical and research [9,12,13].
For the identi cation of valid and reliable instrument, psychometrically valid and reliable culture based instrument is crucial, which helps in the exact assessment of perceived supportive care needs of cancer patients [12,14,13,9].
Many kinds of need assessment instruments are largely developed and validated in English speaking countries for the assessment of supportive care needs of cancer patients. The Supportive Care Needs Survey (SCNS) is a commonly used instrument for the identi cation of perceived supportive care needs of cancer patient [15,1]. SCNS is one of the best, reliable, valid, complete and strong cancer speci c needs assessment instrument it assesses the kind and extent of perceived supportive care needs of cancer patient covering the ve domains (psychological, health system and information, physical and daily living, patient care and support, and sexuality) [16,15].
SCNS-SF34 is a frequently used patient-reported outcome measure (PROM) on generic cancer-related SC needs in which patients give the information regarding present need and degree of support for help in the previous month as a result of having cancer (1-no need, not applicable; 2-no need, satis ed; 3-low need; 4-moderate need; 5-high need). A high score in the tool indicates that perceived supportive care need is high level [6,17].
When PROMs is going to be used in a different setting and different cancer population it may affect the psychometric characteristics so again psychometric validation of PROMs is essential and the importance [14].  [12,20]. The sample size was 340 as per rule of thumb at least 10 subjects per item of the instrument scale along with item analysis and exploratory factor analysis is suggested [21,22]. Data were collected from patients' diagnosed with cervical cancer , undertaking cancer therapy, physically and mentally able to complete the questionnaire and able to understand the Nepali language [23,24] from May 2020 to September 2020.

Process / Procedure
This study was formally approved by the School of Nursing & Health, Zhengzhou University, Henan China and Nepal Health Research Council. The eligible respondents were identi ed by the main researcher and trained survey interviewer nurses (Nurse having Bachelor degree) from in and outpatient departments of selected hospitals. The eligible respondents were informed about the purpose of the study and also assured the standard of care would not modify irrespective whether they participated in the study or not. After getting informed consent from the eligible respondents, they were requested to complete a set of self-report questionnaires on the same day either at the hospital (inpatients) or at home (outpatients). The main researcher and trained survey interviewer nurses checked the questionnaire thoroughly immediately after returning the questionnaire by the respondents to avoid the missing response. Participants who did not return the questionnaire were followed by phone a call.

Measures
Development of the SCNS-SF34-N (Nepali) SCNS-SF 34 is recognized as a valid and reliable need assessment tool for identifying the supportive care needs through a self-reporting questionnaire regarding patients' perceived supportive care needs [25,6,26]. It assesses the existing need and the degree of need for supportive care on the last one month of 34 items by using ve point and two level response scale (1-no need, not applicable; 2-no need, satis ed; 3-low need; 4-moderate need; 5high need). The initial response scale comprises of two broad categories of need, i.e. 'no need' and 'need'. The "no need" scale is additionally subdivided into two categories namely "not applicable and satis ed". Not applicable indicates there were no problems to the patient on the related item and satis ed indicates that for that particular item patient needed support but this support was managed by himself. The "need" scale is additionally subdivided into three categories namely 'low need', 'moderate need' and 'high need' representing the level of supportive care need [6].The 34 items are categorized in 5 domains (psychological needs (10 items), healthcare system and information needs (11 items), physical and daily living needs (5 items), patient care and support needs (5 items) and sexuality needs (3 items). A high score in the tool indicates that perceived supportive care need is high level [6].
Beaton's guideline was used to develop the SCNS-SF34-N (Nepali).it includes the seven scienti c stage namely: 1. Translation into the target language 2. Synthesis of the forward translations 3. Backward translations, 4. Consensus conference, 5. pretest patient survey, 6. Approval of research team 7.
approval of original authors [27,28,9]. Content validity was assessed by consulting 10 experts working and educating in oncology and research area, clarity of the questionnaire was assessed by consulting 15 respondents using Likert scale [12]. Correlation of the questionnaire was assessed employing the test-retest method by consulting 50 respondents [18].
The preliminary nal version was pretested among 34(10% of the total sample) cervical cancer patients in Nepal to determine the understanding level, word appropriateness, identi cation of offensive or aggressive words, identi cation of simple words and acceptability of the translated questionnaire [27] .

Comparative measures for validity testing
To allow comparisons within each item, inter item correlation was assessed for the establishment of convergent validity through Spearman's rank correlation coe cients in between each item of SCNS-SF34-N [29,30,31].

Sociodemographic and clinical characteristics
The sociodemographic characteristics included the data about age, education, marital status, economic status, dietary status, family type, relationship status. The clinical characteristics included disease stage, treatment modalities, duration of disease.

Statistical analysis
Statistical Package for Social Science(SPSS) version 20(IBM, NY, USA) was used for the analysis of collected data. Sociodemographic and clinical characteristics of the respondents were analyzed by using descriptive statistics (frequency, percentage, mean, standard deviation) [9,18].
Exploratory factor analysis (EFA, principal component analysis with varimax rotation) was used for the examination of factor validity. To test the suitability of the data for EFA, The Kaiser-Mayer-Olkin (KMO) and Bartlett's tests were used to identify the sample adequacy and appropriateness of sample size respectively[18].
The suitable parameter of KMO statistic (0.917), exceeded the threshold of 0.5, and Bartlett's test was signi cant (chi2 = 26,958.140, p <0.001), which indicates that the data were suitable for factor analysis [9].
Factor loadings >0.4 and Eigenvalues >1.0 were taken into consideration for acceptance. Items were recognized to the factor with the priority of highest loading and the factor structure was calculated for the explanation of the variance. oor and ceiling effect occurred if more than 50% of the participants attained the lowermost (0) or uppermost (100) score for each factor. [32,19,9]. Cronbach's alpha with a coe cient value >0.7 considered as acceptable which was calculated through internal consistency [32,19,9,14,13].
For the assessment of discriminant validity, independent -sample t test was used to test the differences in the mean score for each domain between numerous subgroups of participants with different sociodemographic and clinical characteristics [19,13,33,9].

Results
Demographics and clinical characteristics of respondents A total of 98.2% (334/340) of eligible participants from selected hospitals and within the selected inclusion criteria provided informed consent and returned the completed questionnaires A total of six eligible participants were missed because they leave hospital the without returning the questionnaire and they did not respond in phone call also.
Among 334 respondents 74.25% of the respondents were below 64 years, and the mean age was: 54.59 years and SD was:12.71. 49.4% of the respondents were in II stage of cervical cancer disease and 63.2% of respondents were on Radiation + Chemotherapy treatment. (Refer to Table 1).  Factor loadings > 0.4 and Eigenvalues > 1.0 were taken into consideration for acceptance oor and ceiling effect were occurred if more than 50% of the participants attained the lowermost (0) or uppermost (100) score for each factor Reliability Internal consistency was achieved at an acceptable level, with Cronbach's alpha coe cients ranging from 0.789 to 0.929for all ve domains and 0.887 for the whole scale (Refer to Table 2).

Convergent validity
For the assessment of convergent validity inter-item, the correlation was assessed. Physical and daily living needs are strongly correlated with psychological (0.599 **) , patient care and support (0.469 ** ) and health system information (0.400 **) . The psychological need is strongly correlated with physical and daily living (0.599 ** ), patient care and support (0.480 ** ) and health system information (0.526 **) . Sexuality need is correlated with physical and daily living need 0(.115 *) .
Health system & information need is strongly correlated with physical (0.400 ** ), psychological (0.526 ** ) and patient care and support (0.643 ** ) needs (Refer to Table 3).  For the analysis of dichotomous variable 't' test was used and as shown in Table 4, illiterate cervical cancer patients reported higher levels of supportive care needs than those with literate cervical cancer patients. Cervical cancer patients not living together with their partners reported a higher level of supportive care need than those cervical cancer patients who are living together with their partners on all domain except physical and daily living.
Respondents who were younger than 65 years and respondents with a life partner reported higher level of sexuality needs than older patients and those without a partner. There was no signi cant difference in the level of supportive care needs either early or late stage of cancer except sexuality domain, respondents in the early stage of cervical cancer reported a higher level of supportive care need in sexuality domain than the respondent in the late stage of cervical cancer. There was no signi cant difference in the level of supportive care needs either patient receiving treatment with radiation or without radiation therapy except in health system information and sexuality domain, respondents receiving treatment with radiation reported a higher level of supportive care needs in health in health system information and sexuality domain than the respondents receiving treatment without radiation (Refer to Table 5).

Discussion
Translation procedure and content validity were separately written in another paper, so here mainly focuses on internal consistency, structural, convergent and discriminant validity.
In this study, more than 95% lled all the 34 items of questionnaire and a missing data rate lower than 5%. Thus, our ndings are similar to the study done in German found that nearly 80% lled all the 34 items of questionnaire and a missing data rate lower than 10% [19].
The SCNS-SF34-N (Nepali) maintained a high level of Internal consistency which was achieved at an acceptable level, with Cronbach's alpha coe cients ranging from 0.789 to 0.929for all ve domains and composite reliability score was 0.887 for the whole scale. These ndings are supported by other studies: Internal consistency was high with Cronbach's alpha coe cients for the ve factors ranging from 0.86 to 0.96 [6], Cronbach's alpha values ranged from 0.82 to 0.94 [19], Cronbach's alpha coe cients ranged from 0.854 to 0.942 for the ve domains and 0.947 for the whole scale [9] and these ndings are also supported by another study in which split half method that was used for the internal consistency with the α value for the rst half is 0.87 while it is 0.92 for the second half, the Spearman-Brown coe cient for the whole of the scale is 0.74 and The Guttman Split-Half coe cient is 0.73 [12].
For the measurement of sampling adequacy, The Kaiser-Meyer-Olkin test was used and the Bartlett test of sphericity test was use to con rm the su ciency of data for conducting explanatory factor analysis. This ndings are consistent with some other studies done in German, China, and Turkey respectively [19,9,12] in which factor structure of the SCNS-SF done by Kaiser-Meyer-Olkin test, Bartlett test of sphericity test and explanatory factor analysis. These ndings are contrast with other study done in China in which the factorial structure of the Chinese version of the SCNS-SF using con rmatory factor analysis in two different colorectal cancer samples, Hong Kong Chinese and Taiwan Chinese patients [13].
Principal component analysis through EFA of the SCNS-SF34-N discovered ve dimensions explaining 65.48% of the total variance that are almost same as the original English version (Psychological, health system and information, physical and daily living, patient care and support, and sexuality needs) German [19], China(Mainland), [9] and Turkey [12]. Findings of our study are contrast with the study Dutch version of the SCNS-SF34 which discovered a four-factor structure, which combined the health systems and information domain and the patient care and support domain into a single domain [14].
The Bartlett's sphericity test that was used to assess the structural validity of the SCNS-SF 34 gave out a value of 8240.401and a level of < 0.001. This result shows that the data set is appropriate for a factor analysis. KMO sample value was 0.889. The fact that the KMO criterion is 0.88 means that the sample size is quite appropriate for factor analysis. These ndings are consistent with the ndings of these study done in German, China and Turkey German [19], China(Mainland), [9] and Turkey [12].  14,18,9]. While in a study done in Turkey, the author had not mentioned about convergent validity [12].
The discriminant validity of SCNS-SF34-N (Nepali) was con rmed by comparison with known subgroups. According to the ndings of this study respondents who were younger than 65 years and respondents with a life partner reported a higher level of sexuality needs than older patients and those without a partner this nding is consistent with other studies [9,13] respondents younger than 60 years reported a higher level of sexuality needs [19]. This study found that respondents in early stage of cervical cancer reported higher level of supportive care need in sexuality domain than the respondent in late stage of cervical cancer and no signi cant difference in the level of supportive care needs an either early or advanced stage of cancer. These ndings are consistent with a previous study [19] except sexuality domain but ndings are contrary to the study done in mainland China found that respondents with advanced disease reported more needs across all domains except sexual needs [9].

Conclusions
The Short-Form Supportive Care Needs Survey Questionnaire in Nepali (SCNS-SF 34N) is a valid and reliable instrument in the context of Nepal in terms of identifying the supportive care needs of cervical cancer patients. The scale contains a total of 34 items. Its validity and reliability have been established via language validity, content validity, internal consistency, structure validity, discriminant validity and Alpha coe cient. Further study is ongoing for the assessment of supportive care needs of cervical cancer patients by using this survey questionnaire (SCNS-SF 34N). We would like to recommend to use this SCNS-SF 34 N for the identi cation of other cancer patients' supportive care needs for the arrange of intervention according to priority basis.