This randomized controlled study involving 102 post-appendicectomy patients indicates no significant effect PEMF has on both the patient-reported pain intensity measures (i.e. sitting and moving) over the initial 12 hours post procedure. Similarly, the secondary endpoint – post-operative use of IV fentanyl via PCA – also did not show any difference in between the intervention arm vs. the placebo arm. Hence, previously reported effect of PEMF on post-operative pain intensity and analgesia use in other surgical settings cannot be verified.
The reason of our negative findings is not immediately obvious. We proposed the following for further deliberation. Firstly, acute post-operative pain occurs as a result of tissue trauma or direct nerve injury.[20] Tissue trauma releases local inflammatory mediators which cause increased sensitivity to stimuli in the area surrounding an injury (i.e. hyperalgesia) or misperception of pain to non-noxious stimuli (allodynia). Additionally, the sensitization of the peripheral pain receptors and increased excitability of central nervous system neurons also contribute to hyperalgesia and allodynia.[20] Different surgical conditions and related procedures might bring about a different mixed of these pain aetiologies, thereby affecting the effect of PEMF on cellular mechanisms that induces pain during the acute post-operative period. Hence the possibility of differing PEMF efficacy in different surgical conditions and procedures should be considered.
Next, it is increasingly known that genetic factors play a significant role in influencing one’s responses to pain and analgesia.[21, 22] For example, our production of opioid metabolizing enzymes and expression of various pain receptors and signal transduction elements are affected by various genetic factors. It also interacts with other physio-pathological, psychological as well as environmental factors that give rise to various pain-related outcomes.[23] These differences are known to give rise to the large variability in terms of pain perception and post-operative pain management needs. Our study was conducted on a sample of south-east Asian who may pose very different genetic make-ups compared to the Caucasian population, where most of the earlier studies was focused on.[19, 24, 25]
Furthermore, existing evidence on post-operative use of PEMF therapy have not been very consistent. Heden and Pilla (2008) reported a controlled trial of PEMF use among 42 females with post breast augmentation surgery.[12] The study indicated a large and significant effect (more than 3 times different) PEMF had on postoperative pain and pain medication use observable from the 3rd day after the procedure. In another setting, Adravanti et al. (2014) assessed the use of PEMF among 33 subjects with total knee arthroplasty. [25] The therapy was administered for four hours a day for a duration of 60 days. The main difference in this study is that the therapy was initiated only after one week of the operative procedure. Subjects were then assessed at one, two- and six-months post-procedure. The study reported that pain score (measured using the VAS) were consistently lower among those applied with PEMF. The differences were consistent across the three postoperative follow-ups. More recently, the interest of PEMF has been extended further to other surgical discipline. A randomized, double-blind, placebo-controlled evaluation of PEMF by Khooshideh et al. (2017) on postoperative pain, analgesia use and wound healing in patients undergoing caesarean section has been reported.[19] This study was motivated by the lack of data on PEMF efficacy for deep organ surgery. Seventy-two women were investigated for their pain intensity (assessed using a VAS) at regular interval starting from 2 hours to 7 days post procedure. Analgesia use were also examined. The data shows that pain score and analgesia use were significantly lower among those assigned to the intervention vs. the placebo control. The effect was again largely consistent across the entire observational period (from 2nd hour post procedure to the 7th day).
Despite these positive results, there are also a significant body of reports with negative findings. For instance, Stocchero’s study in 2014 represented an important piece of evidence with negative findings.[15] In Stocchero’s study, 120 patients undergoing unilateral mandibular third molar extraction were randomly assigned into three study groups (enabled PEMF, disable PEMF and no device) after the procedure. Pain score and analgesia use were assessed for a week. This study, similar to our evaluation, found no significant different in between those applied with placebo or without device and those with PEMF during the seven days post procedure. Similar negative results have also been reported in a number of PEMF studies: post blepharoplasty [26], breast augmentation [24] and dental implantation procedure [14].
A notable feature among the existing evidence on PEMF and its effect on postoperative pain is their small sample sizes in general. Besides the study reported by Stocchero et al. (2014) [15], the study sizes of those described above ranges from 11–72 with the median study size being 36. The problem with a small sample size in clinical research is of two edges. A small study may risk not being able to detect a clinically significant difference due to the lack of statistical power, hence committing a type II error.[27] This is a well-known limitation when a study with small sample size is reported, particularly when the finding is negative. An equally important but less well appreciated problem with small study size is related to a phenomenon known as the “small study effect”.[28] It has been found that, trials with small number of patients are more likely to report larger and significant treatment effects compared to their larger counterparts.[29, 30] One of the possible reason, as illustrated by Kjaergard and colleagues [31] is the lower study quality often associated with small studies. The lack of adequate sequence generating, allocation concealment and blinding can significantly exaggerate the intervention effect compared to large trials. Furthermore, trials with subjective outcomes, such as pain score in this context, also has been reported to have the potential of exaggerated effect sizes when there was inadequate concealment or blinding.[32] This is in contrast to trials with objective outcomes such as mortality, as there was little evidence that inadequate concealment and lack of blinding would distort the estimated effect sizes. Furthermore, the difference in the subjective experience of pain may cause great variations in reported VAS pain scores and need of pain medication. As reported by Svaerdborg and colleagues [24] in their study, patients reported VAS scores of 9 to 10 during the first postoperative day and yet other patients reported VAS scores of only 1. As a result, the risk of registering an underlying difference in pain threshold in the different groups by chance can be potentially large (hence, biased) given a small study size.
Findings from this study should be interpreted with the following limitations. First, different PEMF devices may have slightly different settings. The differences may affect the efficacy and contribute to the inconsistent findings. The potential effect of the different device setting on effectiveness has been previously demonstrated.[13] Our study used a device closely resemble that used previous studies [19, 33] which has obtained a significant positive outcome. Next, our study only collected data that spanned over the 12 hours observation period during the immediate post-operative phase. There is possibility that the effect of PEMF will require longer observation to be noticed. Thus, our conclusion of the lack of effect can only be interpreted within the time limit of our data. It is however worth noting that previously reported positive studies [19] have hinted to a significant effect visible as early as 4 hours post operation. Thirdly, in order to balance between the generalizability of our study conclusion and the study efficiency (and safety), we have excluded the younger as well as the older age patients (i.e. < 19 and > 50 years old) despite their significant representation of the population with appendicitis and appendicectomy. Furthermore, we have also excluded those with complicated appendicectomy (which their operative time has exceeded three hours) and those with potentially altered pain perception (i.e. with underlying chronic pain or long-term use of opioid). These exclusions from the study population has limited the external validity of our study findings and conclusion. Interpretation of our study should therefore take into account of this limited generalizability. Lastly, we excluded some randomized subjects prior to the completion of the data collection because of the various reasons described above. If the dropouts are systematically related to both the treatment assignment and study outcomes, a confounded effect is possible. We lack the data to further analyse these excluded subgroups, hence were unable to inform about the risk of a confounded effect.
In conclusion, this study has not been able to verify the effect of PEMF on postoperative pain previously reported. Given the currently available evidence, a recommendation to using PEMF as a routine part of post-operative pain management is not warranted especially for pain management during the immediate post-operative period and of use for short durations. Future larger scale study may be required to provide more convincing evidence regarding the effect of these devices. A high-quality systematic analysis of the currently available evidence might be required to clarify the current evidence gap in order to better inform future study design and focus.