Background

The number of patients receiving dialysis for end-stage renal disease (ESRD) has been increasing worldwide and projected to more than double by 2030 [1]. Japan has the second highest prevalence of dialysis for ESRD, next to Taiwan, and the second largest number of patients receiving dialysis, next to the USA [1]. The number of Japanese patients receiving maintenance hemodialysis (HD) was over 300,000 in 2012 [2]. Among different types of vascular access available for HD, the native arteriovenous fistula has been historically preferred in Japan, accounting for nearly 90 % [3].

Pain associated with the insertion of dialysis needles can cause considerable stress on HD patients [4]. In fact, pain during needle insertion is the most common patient complaint regarding hemodialysis access [5]. Several approaches have been developed to minimize needle insertion pain, including buttonhole cannulation technique [6, 7], lidocaine-containing local anesthetic cream/patches [8, 9], and lavender essential oil [10, 11]. However, their effects are variable and unpredictable.

Thermosensation and pain sensitivity are closely related. Studies have shown that skin temperatures over 43 °C and below 15 °C evoke a feeling of pain [12, 13]. However, within this range of 15 °C through 43 °C, lower skin temperatures were associated with longer latencies to pain sensation indicating decreased pain sensitivity [14, 15]. Considering these findings, Daimon et al. tried to control pain during needle insertion for HD by decreasing skin temperature to 20 °C with a cooling gel [16]. Skin cooling was found to reduce needle insertion pain significantly and as effectively as lidocaine patches. With no adverse skin reactions related to the procedure, the use of skin cooling was suggested to be more favorable for patients allergic to lidocaine patches.

A handy-sized device named Mutsu-Senshi® (Creative Medical Engineering Co, Ishikawa, Japan) has recently been developed to cool the skin (Fig. 1). In Japanese, “Mutsu” means “painless,” while “Senshi” is a play on words meaning both “centesis” and “fighter (for pain).” Mutsu-Senshi® with a thermoelectric mechanism was shown to bring the skin temperature down to around 21 °C in 1 min, without causing significant changes in the diameter of dialysis vessels [17]. It is expected that Mutsu-Senshi® can be used in the clinical practice more smoothly and safely than the method of using cooling gel.

Fig. 1
figure 1

Presentation of the Mutsu-Senshi® skin cooling device

Full size image

However, Mutsu-Senshi® has never been formally evaluated for its effectiveness in multicenter studies. It is unknown what proportion of patients on HD favor the use of this device in the real-world practice. Another question is if there is any particular group of patients who respond better to Mutsu-Senshi®. This would be useful information for our clinical practice, allowing us to recommend Mutsu-Senshi® for those who will benefit most from this device. Therefore, this multicenter prospective study was planned to assess the effectiveness of the Mutsu-Senshi® skin cooling device in 4 dialysis centers in Japan.

Methods

This study was registered with UMIN Clinical Trials Registry, number UMIN000011802 [18]. After uploading the study protocol, we modified it to simplify the process and encourage a smooth operation of the study in the real work environment. The points changed were as follows: (i) observational periods for each patient were shortened from 3 weeks (a normal week, a week using Mutsu-Senshi®, and another normal week) to 2 weeks (a normal week and a week using Mutsu-Senshi®); (ii) patients were asked to evaluate the effect of Mutsu-Senshi® only at the end of the second week, instead of after every HD session; (iii) the pain scale was changed from a commonly used absolute visual analog scale (VAS) for assessment of pain intensity using numbers (e.g. 0–10) to a modified VAS for assessment of relative changes in pain intensity with Mutsu-Senshi®, as explained later.

Informed consent was obtained from individual patients. The study was approved by the institutional review board of the University of Tokyo.

Study population

We conducted the study in 4 HD centers in Japan (Medical Plaza Shinozaki, Tokyo; Asagaya-Suzuki Clinic, Tokyo; Tokiwakai Group Tokiwa Hospital, Fukushima; Akita City Hospital, Akita; and University of Tokyo Hospital, Tokyo), during the period between January and August, 2014. We recruited patients age 20 to 90 years who were receiving maintenance HD. Exclusion criteria included patients who may overreact to cold stimulation, those with sensory disturbance in the limbs, those with visual disturbance or hearing impairments, those unable to communicate adequately with observers (e.g., dementia), those with needle insertion troubles over the last 3 HD treatments, and those whom responsible doctors consider as inappropriate participants.

Intervention

For the first week after entry into the study, patients received HD in a conventional way, including needle insertion by trained medical staff. For the second week, Mutsu-Senshi® was additionally used, without changing any other procedure conducted in the first week. Patients using lidocaine patches continued to use them in the second week.

First, both the patient skin and the skin contacting surface of the Mutsu-Senshi® were disinfected. After switching on Mutsu-Senshi® and waiting until the temperature of its skin contacting surface dropped to 21 °C and an indicator lamp was lit, the cooling device was put on the access site before dialysis needle insertion. Mutsu-Senshi® is designed to beep either when the skin temperature reaches 21 °C or after 1 min of contact with the skin; 1 min is considered sufficient to decrease the skin temperature to around 21 °C for most users [17]. Finally, dialysis needles were inserted in the same way as the first week.

Assessment

At the end of the second week, we asked each participant to assess the effect of Mutsu-Senshi®. A modified VAS was created for this purpose, scoring from −100 indicating a marked increase in pain intensity with Mutsu-Senshi® to +100 indicating a marked reduction in pain intensity with Mutsu-Senshi®. The score ±0 indicates that patients did not find any difference with or without Mutsu-Senshi®. The participants were instructed about the modified VAS score entry on a tablet computer. No medical staff observed the process of entry to reduce response bias. The results were transferred electronically to the Lifestyle Disease Coordinator Association in Japan, anonymized, and then sent to the authors for analysis in the end of the study. The Lifestyle Disease Coordinator Association is totally independent of the manufacturer of Mutsu-Senshi® as well as participating HD centers.

Statistical analysis

Descriptive statistics were gathered for all eligible patients. Age was shown as mean and standard deviation (SD), while HD vintage was expressed as median and interquartile range (IQR). Categorical variables were shown with number of patients and proportion. Distribution of modified VAS was presented using a scatter plot as well as a box plot with the scores of median, IQR, minimum, and maximum. Scores outside of 1.5 times IQR were regarded as outliers and thus excluded for the box plotting. Then, subgroup analysis was conducted to identify any particular patient group benefiting more from Mutsu-Senshi®. Patients were classified into 2 groups for the following factors: age (<65 or ≥65), sex, HD vintage (<5 or ≥5 years), reason for ESRD (non-diabetes or diabetes), needle insertion site (forearm or upper arm), the use of lidocaine patches, and diameter of dialysis needles (15–17 gauge or 18 gauge). The cut-off points for classification were determined such that the number of patients in each subgroup was relatively similar. Box plots were used for the individual subgroup, and then a Wilcoxon rank sum test was conducted to examine whether any predefined factor was associated with increased preference for Mutsu-Senshi®. Finally, a multivariate logistic analysis was conducted. The outcome was modified VAS score >0 (vs. score ≤0), while exposures were all the variables used for the aforementioned subgroup analysis. All data were analyzed using Stata 14 software (Stata Corp, TX, USA). A P value of <0.05 was considered statistically significant.

Results

Ninety-eight patients (mean age 60.5 [SD 13.8], men 66 %) participated in the study (Table 1). All patients had native arteriovenous fistulas. All of them completed pain assessment and reported no adverse effects associated with Mutsu-Senshi®. As shown in Fig. 2, modified VAS scores were scattered from −100 to +100, and the box plot demonstrated the median score of +29 and IQR from ±0 to +58. The proportion of patients scoring >0 (i.e., favoring Mutsu-Senshi®) was 64 % (63/98). In a subgroup analysis (Fig. 3), there was no clear difference in the preference for Mutsu-Senshi® between younger (<65) and older (≥65) patients, between men and women, between shorter (<5 years) and longer (≥5 years) HD vintages, between patients for whom diabetes was or was not a cause of ESRD, and between the forearm and upper arm for needle insertion. However, patients using lidocaine patches tended to favor Mutsu-Senshi® compared to those not using them (P = 0.089). Patients using 15–17-gauge needles demonstrated significantly stronger preference for Mutsu-Senshi® than those using 18-gauge needles (P = 0.006). In a multivariate logistic regression analysis (Table 2), needles with larger diameters (15–17 gauge vs. 18 gauge) were significantly associated with favoring Mutsu-Senshi® (adjusted odd ratio 5.69, 95 % confidence interval [CI] 2.17–14.96). The patients were further divided into 3 groups by needle diameter (3 patients using 15 gauge were grouped into those using 16 gauge): 15–16 gauge (n = 31), 17 gauge (n = 27), and 18 gauge (n = 40) in a multivariate regression model. Compared with patients using 18-gauge needles, the adjusted odds ratio in favor of Mutsu-Senshi® was 4.96 (95 % CI, 1.38–17.76) for those using 17-gauge needles and 6.25 (95 % CI, 2.01–19.41) for those using 15- or 16-gauge needles. These findings mean that, although there was a significant difference between 15–17-gauge and 18-gauge needles, no significant difference was observed between 15–16-gauge and 17-gauge needles.

Table 1 Baseline characteristics of study participants
Full size table
Fig. 2
figure 2

Results of the modified visual analog scale presented as scatter and box plots

Full size image
Fig. 3
figure 3

Subgroup analyses for the assessment of Mutsu-Senshi® by factor and P value estimated by Wilcoxon rank sum test. ESRD end-stage renal disease. *Median value was ±0 (the same as the lower side of an interquartile range)

Full size image
Table 2 Multivariate logistic regression analyses for favoring Mutsu-Senshi® (modified visual analog scale >0)
Full size table

Discussion

In this first prospective multicenter study, the Mutsu-Senshi® skin cooling device was suggested to be effective in reducing needle insertion pain in patients receiving HD. Mutsu-Senshi® was safely used in all of the 98 patients without any complications, and 63 (64 %) preferred Mutsu-Senshi® to conventional needle insertion without skin cooling. We also found that those using needles with larger diameter tended to rate Mutsu-Senshi® more favorably, and this factor was significant even in the multivariate regression analysis.

Needle insertion pain is a considerable stress for patients receiving HD [4, 5]. It may be associated with depressive symptoms and lower quality of life in patients with HD [19]. Therefore, various methods have been attempted to reduce needle insertion pain [6–11]. The buttonhole technique has been expected to decrease pain more effectively than the rope-ladder or rotating-site cannulation technique [6], but a recent meta-analysis concluded that no difference in cannulation pain was found between the 2 techniques in randomized controlled trials [7]. In contrast, lidocaine patches have been shown to be effective in reducing pain [8, 9] and were actually a popular pain management modality used by 40 of the 98 patients in our study population. However, lidocaine patches frequently cause skin problems such as pruritus and eczema. A Japanese study reported that among 26 HD patients using lidocaine patches, 15 complained of skin problems [20]. In our study, none of the participants reported skin problems associated with the use of Mutsu-Senshi®. This is probably because the extent of skin cooling with Mutsu-Senshi® was mild, only to 21 °C.

Sixty-four percent (63/98) of patients rated Mutsu-Senshi® more favorably for pain relief, although the median value of the modified VAS was not very large at +29 (IQR from ±0 to +58). A univariate analysis suggested that a subgroup of patients using lidocaine patches tended to favor Mutsu-Senshi®. We speculate that patients using lidocaine patches may be more sensitive to pain, or more particular about pain control methods, and therefore evaluated Mutsu-Senshi® more favorably than those not using lidocaine patches. Our findings also suggest that Mutsu-Senshi® can be used in combination with lidocaine patches in the real-world practice. In both univariate and multivariate analyses, larger needle diameters were associated with a higher preference for Mutsu-Senshi®. A potential explanation would be that patients using needles with larger diameters (15–17 gauge) might have had stronger needle fear. Therefore, the use of Mutsu-Senshi® might have provided them with greater relief than those using smaller diameter (18 gauge) needles. However, no difference was observed between 15–16-gauge and 17-gauge needles, which may be explained by the small sample size or influence of confounding factors. Further study may be necessary to examine the association between needle diameter and the usefulness of Mutsu-Senshi®.

This study has several strengths. First, patients were recruited from 4 HD centers, increasing the generalizability of the results. Second, all the prescribed conditions were kept unchanged between the first and second weeks, except for the additional use of Mutsu-Senshi®. In particular, patients using lidocaine patches were allowed to continue this use through the second week. Therefore, our findings on the usefulness of Mutsu-Senshi® can be quickly applied to the current bedside practice without any modification. Finally, the modified VAS was administered while the medical staff were not observing. Patients may respond to the questionnaire differently if they know they are being watched, probably leading to a more favorable assessment of the new treatment. By reducing response bias, we fairly examined the usefulness of Mutsu-Senshi®.

However, this study also has limitations. First, an ideal way of assessing the effect of Mutsu-Senshi® may be to ask patients about their pain at the end of every HD session, because pain severity can vary from day to day. Calculating and comparing the average absolute pain scores of several needle insertions with and without Mutsu-Senshi® could have achieved even fairer assessment of Mutsu-Senshi®. However, to recruit many patients and operate the study in the real-world clinical environment with a limited number of medical staff, we had to change the protocol to a single assessment comparing needle insertion pain with and without Mutsu-Senshi® only 1 time at the end of the second week. For this purpose, we had to modify a commonly used absolute VAS to a comparative one, making it difficult to compare the effectiveness of Mutsu-Senshi® with other pain control modalities such as lidocaine patches [8, 9] and lavender essential oil [10, 11]. Second, our study population mostly included patients receiving HD for 1 year or longer. Therefore, the study participants might be more tolerant of needle insertion pain than those recently introduced to HD. It remains unknown whether Mutsu-Senshi® is beneficial for patients who have undergone HD for a shorter time period and who may be more sensitive to the pain. Finally, although the diameter of dialysis needles was suggested to be an independent factor for increased preference for Mutsu-Senshi®, this finding might be influenced by unmeasured confounding factors, such as skin thickness and mental health conditions. Further studies are warranted to examine the usefulness of Mutsu-Senshi® and to identify the most appropriate target population.

Conclusions

This multicenter prospective study suggested that the Mutsu-Senshi® skin cooling device may help HD patients with native arteriovenous fistula to better cope with needle insertion pain without causing skin complications. The majority (64 %) of HD patients favored the use of Mutsu-Senshi® before needle insertion. Patients using needles with larger diameters may benefit more from Mutsu-Senshi®.