The impact of pain on the quality of life of patients with end-stage renal disease undergoing hemodialysis: a multicenter cross-sectional study from Palestine

This was a multicenter, cross-sectional study that was conducted between August and November of 2018 and occurred in the hemodialysis units of five different hospitals: Tulkarem Governmental Hospital, Qalqilya Governmental Hospital, Jenin Governmental Hospital, Tubas Governmental Hospital, and An-Najah National University Hospital (NNUH).

For this study, we included five hemodialysis centers that serve approximately 570 patients in the Northern West Bank [6]. We used the Sample Size Calculator by Raosoft Inc. (http://​www.​raosoft.​com/​samplesize.​html), with 5% margin of error, 95% confidence interval, a population size of 570. The calculated minimal required sample size was 220. Then we recruited a total of 330 patients by convenient sampling as follows: 74 from Tulkarem Governmental Hospital, 36 from Qalqilya Governmental Hospital, 88 from Jenin Governmental Hospital, 24 from Tubas Governmental Hospital, and 108 NNUH. HD patients were chosen from each kidney dialysis unit using the proportional quota sampling approach to represent the general HD population [21]. The number of patients from each hospital was determined based on the number of patients undergoing dialysis in those dialysis unit. The rationale for this choice was to make the percentage of patients included in each hospital representative of the percentage of patients undergoing dialysis in that hospital from the total number of patients served in all included hospitals. Eventually, participants were selected using a nonprobability convenience sampling technique because of the associated it with saving researchers time, and costs. It was also the easiest method for the accessibility and resources available for the study [22].

The inclusion criteria were that a patient must be aged ≥ 18 years and have been on HD for ≥ 6 months. We excluded any patient who refused to give consent or was unable to proceed with the interview to the end.

The patients were evaluated by administering a questionnaire consisting of 4 sections. In the first section, we enquired about demographic and social characteristics including age, gender, occupation, educational level, residency, monthly income level, smoking status, and body mass index (BMI), which was calculated by dividing the subject’s weight in kilograms by their squared height in meters squared. For analysis purposes, smoking status was classified into nonsmoker, light smokers (1–9 cigarettes per day), and smokers (≥ 10 cigarettes per day) [23], whereas BMI was classified into obese weight range (BMI of ≥ 30), overweight range (BMI between 25 and 29.9), normal weight range (BMI between 18.5 and 24.9), and underweight range (BMI of < 18.5) [24].

The second section contained items on clinical and HD-related data, including dialysis vintage (length of time on dialysis treatment in years), frequency per week, duration of sessions, transplantation history (for patients who returned to dialysis after rejection of transplantation), and the number of chronic comorbid illnesses and medications taken chronically.

In the third section, we included the Brief Pain Inventory (BPI) scale, a widely used and accepted pain assessment scale that addresses pain severity, as well as pain interference with daily functions [25]. We obtained permission to use the Arabic BPI-Short Form version of the MD Anderson Cancer Center, which has been previously translated and validated [25, 26]. A Lebanese study found that when the BPI scale was used in a study of Arab patients experiencing pain, the Arabic-language translation showed cultural sensitivity and adequate psychometric properties comparable to the original version [27]. The following topics are assessed in the items on the severity of pain: the worst pain felt during the last 24 h, the least pain felt during the last 24 h, the average pain felt during the last 24 h, and the pain felt at the moment of assessment. Topics covered in the items on the interference of pain with daily functions include: the ability to walk, general mood, ability to work, sleeping quality, social relationships, and the ability to enjoy life generally. BPI also contains items addressing pain sites, pain response to analgesics, and the degree of this response. Responses to the four items on pain severity received a score in the form of a number of points between 0 and 10 for each item, and then a pain severity score was created for each participant by adding up the points received on all four items. Accordingly, pain severity scores ranged from 0 to 40. Likewise, responses on the seven items on pain interference received a score that ranged between 0 and 10 for each item, and a pain interference score was created by adding up the points achieved on all seven items, making the range for this score 0 to 70.

In the fourth and final section of the questionnaire, we included the European Quality of Life Scale 5 dimensions (EQ-5D) assessment tool, which is widely used to evaluate the overall quality of life [28]. EQ-5D is one of the most widely used standardized HRQOL questionnaires available today and has been used in many disorders, including hemodialysis, for clinical assessment in health care [29, 30]. The EQ-5D contains five items that cover the following five domains: self-care, mobility, daily activities, pain and/or discomfort, and anxiety and/or depression [28]. It also contains a visual analog scale—the European Quality of Life visual analogue scale (EQ-VAS)—that allows subjects to report on their perception of their health and life quality by answering items on a grading system that ranges from 0 to 100, with 0 representing the worst possible health situation and 100 representing the best possible health situation [31]. We obtained approval to use the Arabic version of EQ-5D in this study from the Euro-QOL Research Foundation (registered ID: 28409). The principal investigator in the current study has previously published a guide on the use of the EQ-5D scale, as well as other projects that employed this tool [32‐34]. The EQ-5D score indexing that was used here has been illustrated elsewhere [34‐38] and used the EQ-5D-5L Crosswalk Index Calculator [39] based on the scoring algorithm of the UK general population. Two academic experts specialized in clinical pharmacy, who also have expertise in research on QOL and statistical analysis, reviewed the questionnaire components in terms of face validity and content validity, as well as clinical accuracy. Face-to-face interviews were implemented by trained researchers for data collection. The questionnaire was pilot-tested on 15 subjects, which were not included in the final study, to determine its comprehensibility and clarity, and the time needed to complete it.

A total of 330 ESRD patients undergoing HD were contacted to participate in this study; 300 of them decided to participate, whereas the rest refused to take part in this study, accounting for a response rate of 90.9 percent. Demographic and clinical variables are presented in detail in Table 1. The participants’ mean age in years was 54 ± 16, with 60.7% of them being < 60 years old. Slightly more than half of the participants (55.3%) were males, and 73.3% of all subjects were married. Subjects living in villages accounted for 58.3% of the sample, and those living with their families accounted for 93.7%. The majority of participants (59.3%) lived in a household with an income of < 2000 NIS (1 NIS = 0.31 US Dollars) per month.

Regarding HD-related characteristics, 65.3% had been on HD for less than four years, 92.0% received HD three times per week, and 86.0% received HD sessions for a duration of less than four hours per session. Only 4.3% reported a history of kidney transplantation. Most subjects had less than three other chronic illnesses, with only 35.7% living with ≥ 3 chronic comorbid diseases. However, 96.7% were on ≥ 4 chronically used medications and 82% of all subjects reported taking their medication by themselves.

The means for pain severity score and pain interference score as found in our study were 10.55 ± 10.62 and 19.41 ± 18.51, respectively, whereas the medians for these scores were 8 [0.00–17.75] and 16.5 [0.00–30], respectively. Additionally, the interreliability indices for the pain severity and pain interference subscales were good, with a Cronbach’s alpha of 0.947 and 0.963, respectively.

Our sample's median EQ-5D score was 0.68, with an interquartile range of [0.51–0.84], whereas the mean EQ-5D was 0.63 with a standard deviation of 0.34. Internal consistency was measured by Cronbach’s alpha, and the result was 0.936. On the other hand, the median EQ-VAS score was 60.00 [60.00–80.00], whereas the mean EQ-VAS score was 57.53 ± 25.09. We found that the distribution of patients who reported the worst health status across all dimensions of EQ-5D was as follows: mobility 3.3%, usual activity 3%, self-care 4%, pain and/or discomfort 2%, and depression and/or anxiety 3.3%, whereas the percentage of patients who reported the worst in any dimension was 2% (Fig. 1).

Table 1 shows the significant differences (p < 0.05) between participants according to age, gender, BMI, educational level, employment, monthly income, dialysis vintage, history of kidney transplantation, chronic medication use, and smoking history in association with EQ-5D scores. Table 1 presents these results in detail. We also found significant negative correlations for both pain interference score (r of -0.868, p < 0.001) and pain severity score (r of -0.783, p < 0.001) with the EQ-5D score.

Table 2 shows the results of the relationship between participants’ characteristics and their EQ-VAS scores. Age, BMI status, educational level, work, monthly income level, dialysis vintage, chronic diseases, and chronic medications were all significantly correlated with EQ-VAS score. We also found a significant negative association between EQ-VAS score and both pain severity score (r = − 0.590, p < 0.001) and pain interference score (r = − 0.647, p < 0.001), as well as a significant positive association between the EQ-5D score and EQ-VAS score (r = 0.742, p < 0.001). The results of the linear regression analysis showed that patients aged < 60 years (p < 0.001), with lower pain severity scores (p = 0.003), or with lower pain interference scores (p < 0.001) had significantly higher QOL scores. Table 3 summarizes the results of multiple linear regression analysis of factors associated with QOL.

Our study’s strengths include its multicenter setting and its relatively large sample size. This was also the first study in the West Bank, Palestine, to examine the effect of pain and other social, demographic, and clinical characteristics on QOL among ESRD patients on HD. Furthermore, the current study used the EQ-5D and BPI scales, which are widely accepted assessment tools to study QOL and pain symptoms, respectively. However, this study had some limitations, including its cross-sectional design, which left any speculation on causality unreliable. Moreover, using the convenience sampling technique may have reduced the generalizability of the results of this study to other HD patients. Finally, even though the scale we used to assess QOL is a widely used and accepted tool, using other scales in addition to or instead of it might have yielded additional data that would have helped us better characterize QOL in our target population.