Reliability and Validity of the Quality of Life – Family Version ( QOL-FV ) in Turkish Family Caregivers of Patients with Cancer

Cancer affects the individuals diagnosed with cancer and their families physically, psychologically and socially in a negative way, and this leads to heavy costs on the part of the family and society. The effects of cancer on the individuals and their family have increased gradually in time and these negative changes have affected the way of life and expectations of the patient as well as the whole family; and therefore threaten the quality of life of both


Introduction
Cancer affects the individuals diagnosed with cancer and their families physically, psychologically and socially in a negative way, and this leads to heavy costs on the part of the family and society. The effects of cancer on the individuals and their family have increased gradually in time and these negative changes have affected the way of life and expectations of the patient as well as the whole family; and therefore threaten the quality of life of both 2009; Pereira, 2011).
The family caregivers were under stress to a large extent and had an increasing risk in terms of suffering from depression and having other health problems and McMillan et al., 2005). In some studies, it is shown that family caregivers of cancer patients have experienced more anxiety and depression (Couper et al., 2006). Primary caregivers could be suffering from concurrent emotional distress due to the extensive demands associated with the

Reliability and Validity of the Quality of Life -Family Version (QOL-FV) in Turkish Family Caregivers of Patients with Cancer Figen Okcin 1 *, Ayfer Karadakovan 2
giving of care to patients with a disease at an advanced stage. Several emotional reactions to caring have been noticed in carers, such as feelings of fear, uncertainty, hopelessness and mood disturbances (Costa-Requena et al., 2011;Pereira, 2011) The involvement and support of the family in every stage of care are of great importance as the family caregivers and the patients are both affected in the same cancer survivors experience lasting effects of disease or its treatment leading to health and employment limitations. In during time of disease, caregivers play an important personal care support (Yabroof andYoungmee, 2009). The family members giving care for the patients with ovarian cancer felt themselves very tired and they had many health problems since the initial period of the disease (Ferrell et al., 2002). In the same study, the relatives of the patients mentioned that they felt themselves quite nervous and quit their jobs in order to provide care for their patients.
To our knowledge, there have not been enough studies on quality of life of family caregivers with cancer patients.
instrumentation study conducted to estimate the reliability reliable and valid instrument in Turkish will help nurses to assess information and support needs of family caregivers with cancer.

Materials and Methods
The study included 218 people whose family caregivers had cancer presenting to the chemotherapy unit 2006 and September 2006. The 218 participants met the All participants gave informed consent.
Inclusion criteria for family caregivers were as follows: 1) not diagnosed with any kind of cancer, 2) 18 years old or older, 3) ability to read and write in Turkish.
: Sociodemographic data included the participant's age, education level, current marital status, work status and patients of relationship with family caregivers. : The scale developed by Ferrell and Grant for assessing the quality of life of cancer patients was reviewed to evaluate the life quality of family members and was re-adapted. The scoring should be based on a scale of 0=worst outcome to 10=best outcome. Several items have reverse anchors and therefore when you code the items you will need to reverse the scores of those items. For example, if a subject circles "3" on such an item, (10-3=7) thus you would record a score of 7. The items to be reversed are 1-4, 6, 13-20, 22, 24-29 and 33. Subscales can be created for analysis purposes by adding all of the items within a subscale and creating a mean score. and tested from 1994-1998 in a study of 219 family caregivers of cancer patients. The test-retest reliability was r=0.89 and internal consistency was alpha r=0.69. Factor the instrument (Ferrell and Grant, 2005).
Multidimentional Scale of Perceived Social Support (MSPSS), which is a kind of scale, developed by Zimet et al. (2008), is composed of 12 items which subjectively different sources. The structure of the subdimension proposed includes the support of family, friend and someone special. The internal consistency of the scale and MSPSS was 0.89.
Measures to ensure language equivalence studies, the most practiced method of back translation method faculty members and two foreign language specialist, translated into Turkish. Translations were reviewed by of oncology and nursing faculty members were evaluated by a group of 10 people consisting of faculty members. Latest version as given scale pilot study was performed to a group of 15 people. Appropriateness of expressions are evaluated. on demographic questionnaire, and cancer-related variables (e.g. type of cancer, metastasis and duration of by the researcher through face to face interview with the minutes.
Before we assessed the reliability and validity of the questionnaire for the Turkish family caregivers, we obtained approval from Betty Ferrell, who developed the Arkar, who adopted to MSPSS to Turkish. We received written approval from the Ethics Committee of Ege of our study to the eligible family caregivers and sought their verbal informed consent for study participation.
consistency of the scale was tested with the item-to-total Spearman-Brown and Gutmann Split Half correlations to estimate the stability of the Turkish version of the analyses were made.

Results
The family caregivers were between the ages of 46-56 (32.6%), a great number of them were male (52.8%) and a great number of them (64.7%) were the husbands or wives of the patients. A number of participants (35.8%) were university graduate, a great number of participants were married (92.2%) and were unemployed (56.4%). A great number of participiants weren't have a chronic disease (67.4) ( Table 1).
The original questionnaire was translated to Turkish by the investigators, whose native language is Turkish. Subsequent translations of the questionnaire were made by 4 people who have a good command of both Turkish and English. The most suitable expressions were selected from the translated versions of the questionnaire, and a  single version of the questionnaire was obtained. Back translation of the questionnaire into English was done by 1 person whose native language was English, who had a good command of both languages and who did not see the orijinal English version of the questionnaire. The back-translated version was compared with the original questionnaire. The questionnaire whose linguistic validity 15 family caregivers with cancer patients presented to the Turkey, and necessary changes in the questionnaire were made in accordance with recommendations from the family caregivers.
offer their opinions about the questionnaire whose linguistic was scored on a 10-point scale: 1 corresponding to the worst and 10 corresponding to the best. Statistical analyses that the experts were in agreement (W=0.154, P=0.061).     The questionnaire on which the experts agreed was piloted for 15 family caregivers to the chemotherapy unit. The family caregivers stated that the items of the questionnaire were understandable and, they did not make any recommendations.
As a result of the factor analysis was collected under the original structure of the 4 subdimension is directed to ensure compliance. In view of the factor weights, it was observed that four factors were haronic with the original scale. Factors for 47.45% of the total variances revealed.
as the factors revealed 16. 25, 12.82, 10.74, 7.62. To ensure compliance with the conceptual framework of the original structure of the scale and dimensions of factors from the difference between subscales, factor loadings 0.10 is less than the appropriate change in the items below can be are between 0.27 and 0.83 (Table 4). Items were named under four different groups in the subscales of the scale which was restructured. According to this, the conditions of psychological and spiritual wellbeing (11 items), psysical health (9 items), the way of approach to diagnosis (7 items), support and economic effect (4 items) ( Table 5).
The Pearson moment multiplication correlation among the scales in a positive direction (r=0.29) ( Table  6).

Discussion
After the linguistic validity of the questionnaire was experts to offer their opinions about the questionnaire whose linguistic validity was confirmed. Statistical showed that the experts were in agreement (W=0.154, P=0.061). The questionnaire on which the experts agreed was piloted for 15 family caregivers to the chemotherapy unit. The family caregivers stated that the items of the questionnaire were understandable and, they did not make any recommendations.
Test-retest reliability varies based on the feature with the measured time interval between applications is usually between 2-6 weeks is adequate. To demonstrate the stability against time to determine the correlation coefficient calculated is expected to be positive and high. This value the scale was calculated as r=0.86 respectively. In a study conducted by the test-retest correlation of the scale was 0.82 (Çimen et al., 2005). Reliability and validity of the The Multidimensional Relationship Questionnaire (MRQ) in Turkey, it was conducted by the test-retest correlation Item analyses require that correlations between scores for each item and the total score on the questionnaire should be made to determine to what extent each item is associated with the whole measurement tool. The item-tomore than 0.25. However, the criterion used in practice However, this is not obligatory. It is recommended that items with low correlations should be deleted. Low itemfrom 0.00 to 0.64 for items. Six items (12,14,22,31,32,34), which were threatened the reliability in the scale and, whose total scale correlation values were under 0.20, were eliminated from the scale. between 0.60 and 0.80, then it will be adequate enough to be reliability for the total scale was 0.90 and subscales was calculated as 0.83 and the Gutmann split half and validity study of the scale of the multi-dimensional (0.89 for Helplessness, 0.87 for Rumination and 0.56 Cancer Fatigue Scale validity and reliability study, Reliability and validity study of the Hospital Anxiety and Depression Scale anxiety subscales of the scale for the two half-test reliability was calculated as 0.69. In another Factor in the calculation of eigenvalues (eigen values) are used, the eigenvalue is calculated as the number of factors. The simplest method is known eigenvalues at 1.00 in the values taken on. A good factor in the transformation, loads, it was seen that the most appropriate method was found in the original scale and in 4 factors. Factors for 47.45% of the total variances revealed. The variance revealed 16.25, 12.82, 10.74, 7.62. Factor loadings of items are between 0.27-0.83.
The scale of some items into more than one factor has on the factors which items enter the criteria to be taken the original structure of the scale taking into account the suitability of items; factor groups where four of the items to the distribution of subscales was changed names. The (11 items)", the second subscale "Physical health status (9 items)", the third subscale "Diagnostic approach to the work, a technique used frequently to test hypotheses. In this method, the investigator, or observations about the resources, in line with the assumption that the relationship already established between them and the level of the direction of the relationship is tested by evaluating the et al. (2004) in their study of depression and anxiety levels have shown that caregivers about six times as much among the scales in a positive direction (r=0.29). Subscale relationship between subscale scores were positively and