Statistics from Altmetric.com
We are delighted to answer Nielsen et al's1 comments that our case report exaggerated the risk of cupping therapies, which was not, in fact, the purpose of our report.
According to the patient's report, bloody discharges from blood blisters on most cupping sites were observed during most sessions. This is not the transitory ecchymosis seen with ordinary dry cupping therapy, but presumably a consequences of malpractice. It should again be noted that dry cupping without needle penetration produced such bleeding, which clearly contradicts the definition of dry cupping itself. Colonoscopy at the time of admission for anaemia correction revealed no lesion to cause the anaemia. No past history of anaemia and no relevant family history (eg, autoimmune disorders and bone marrow disease) were identified by the clinical history. We admit this information should have been addressed in our previous case report,2 and agree with the need for thorough investigation and reporting to exclude any underlying causes of the anaemia besides the cupping treatment.
Although it is logical to exclude all underlying conditions to suggest causal association between interventions and observed adverse events, this is not always possible. To make the best inference with limited information, detailed descriptions of the clinical context may complement the inevitable incompleteness of existing records for patients’ conditions, and allow readers to check the validity of intervention and the delivery process to determine the plausibility of such inferences. In this sense, we argue that the relatively more in-depth description of interventions and clinical context is a major difference between our report and that of Yun et al3 (table 1).
Additionally, we would like to respond to the critiques by Nielson et al. First, we admit that the term ‘resulted from’ in the abstract may exaggerate the causal relationship between cupping therapies and observed adverse events. However, such descriptions were avoided in the rest of our report, and adverse events were illustrated as phenomena occurring after cupping treatments, which do not necessarily suggest a causal relationship. Second, both conventional and traditional medicine plays a role in the management of chronic conditions in South Korea's pluralistic medical system.4 Therefore, the history of poor response to conventional management should not imply that all treatment options had been attempted. Third, our case report addressed the adverse event after cupping by an unqualified practitioner, rather than the safety of cupping therapies performed by competent hands.
Meanwhile, the high risk of attrition and reporting bias of included Chinese studies should be considered when interpreting the few adverse events in the cited systematic review.5 The possibility of the misclassification of worse health outcomes for simple withdrawal or selective reporting of adverse events, such as risk of under-reporting adverse events in systematic reviews of drug trials,6 should not be excluded. Limited generalisability of the safety of cupping therapies in a few individual studies should be addressed by the accumulation of more sound large-scale, pragmatic, controlled, and observational studies.
We hope these discussions will provoke ideas for future research on the safety of cupping therapies.
Competing interests None.
Provenance and peer review Commissioned; internally peer reviewed.