Escribe una PREreview

Changes in v1.2.4, v1.2.5 and v1.2.6 (20–22 September 2026). Three revisions since the deposited v1.2.3. None alters a value, a table entry, a figure datum, an evidence grading or a conclusion; every one corrects a statement the manuscript made about its own method or its own bibliography.

v1.2.4. Two wound-field values in §4.3 were quoted as “122 ± 9” and “59 ± 5” mV/mm without stating that the intervals are standard errors of the mean or giving the number of wounds. Both now carry SEM and n, and the source’s intermediate value (79 ± 4 mV/mm, SEM, n = 20) is added. Table 3’s pressure-ulcer row now records that eight of the 20 trials in that Cochrane review were manufacturer-funded.

v1.2.5. The pre-submission condition — an undertaking that the author would read every source relied on for a quantitative claim before submission — is replaced by a description of the verification actually performed. Thirty-three of the sixty-five such sources are behind paywalls, so the undertaking was one the manuscript could not honestly report having met; the replacement states plainly that not every source was read in full text. This is a weaker claim and a true one. The count of quantitative sources corrects from 63 to 65, and aim (ii) and Table 2’s caption now state the exception the table always carried. The Table 3 harms audit was also completed across all nineteen rows: eight required correction, eleven were compliant, and no row is now unverified.

v1.2.6. Two corrections to the Declarations, found by comparing this manuscript’s declarations against those of its two companion manuscripts rather than by reading it alone. The Use of AI-assisted tools statement named five activities and omitted a sixth — the drawing of Figures 1–4, which were produced programmatically in Python and Matplotlib, contain no generated or photographic imagery, and whose submitted files carry embedded C2PA Content Credentials recording that provenance. The identical omission had been corrected in a companion manuscript the previous day and had not travelled here. The Availability of data and materials statement now also records that four of the 145 numbered references carry no indexed identifier — a book, two IEC standards and a guideline — a fact §11 stated and the declaration did not.

Change in v1.2.3 (19 September 2026). The Availability-of-data statement is corrected, and nothing else is. It said the deposited database contains 175 records, and it quoted a version DOI alongside the concept DOI. Both had gone stale: the database now holds 210 records, and the version DOI quoted — 10.5281/zenodo.22831176 — resolves to a superseded version that was missing 5 of the sources cited by one companion manuscript and 29 of those cited by another. A reader following that identifier would have landed on the incomplete deposit. The statement now quotes the concept DOI alone, which always resolves to the current version, and says why no version DOI is given. It also names the four references here that have no database row, because they are a book, two IEC standards and a guideline. No scientific content, value, table, figure or reference changed.

Change in v1.2.2. The reference database promised in the Availability-of-data statement now exists and is cited by DOI: 10.5281/zenodo.22831175 (concept), registered on Zenodo as a supplement to this manuscript. It held 175 records with a crosswalk column mapping each to its numbered reference here, a codebook, the search log, and the script that reproduces Section 5.2. One figure in Section 5.2 (the field in fat at 1 mA) is corrected from 38.5 to 38 mV/mm, so that all six figures are given to the two significant figures the section states it uses. No other scientific content, value, table, figure or reference changed.

Change in v1.2.1 (file only). The Availability-of-data statement in v1.2 carried an unresolved placeholder, "[DOI to be added]". It is replaced with a statement of fact: the verified reference database will be deposited before submission and linked from this record, and no identifier is quoted for a deposit that does not yet exist. No scientific content, value, table, figure or reference changed between v1.2 and v1.2.1.

Preprint. Not peer reviewed. Not submitted to any journal.

Electrical quantities are set at the terminals of bioelectronic devices, yet tissue responds to quantities that are rarely reported: the electric field at the cell, current density, charge per phase and the time course of membrane polarisation. This narrative review assembles, from primary literature and systematic reviews, a framework linking voltage, current, frequency, waveform and charge to their biological counterparts. It covers membrane biophysics; measured electrical properties of human tissues; endogenous fields in development, regeneration and wound repair; electrophysiological measurement; and the transfer problem by which surface-applied current is attenuated and redistributed before reaching a target. It examines what frequency does and does not explain, benchmarks established electrotherapies against their best available evidence, cautions against cross-species inference, and proposes a minimum parameter reporting set and a ten-category scheme for classifying claims by evidential status.

Novelty has not been independently established for either proposed instrument. No formal prior-art or scoping search was performed; the related prior work (Peterchev et al. 2012 on transcranial dose reporting; TIDieR and the CONSORT extension for non-pharmacological treatments; SENIAM) is named in Section 9.1.

Version 1.3.0 (29 September 2026). This version REMOVES material, and that is the whole of it. Section 9 carried two instruments - a minimum parameter reporting set and a ten-category evidence classification - as its Tables 4 and 5, drawn as its Figures 3 and 4. Both are now developed in full in a companion paper by the same author (concept DOI 10.5281/zenodo.23043340; version 0.3 DOI 10.5281/zenodo.23043341), and carrying them here as well would be redundant publication. Sections 9.1 and 9.2 are reduced to outlines citing it; Tables 4 and 5 and Figures 3 and 4 are removed; and the remaining tables and figures are renumbered, so that the table previously numbered 6 - this review's own claims, classified - is now Table 4. Readers comparing versions should note the renumbering: Table 4 of v1.2.6 is not Table 4 of this version. Section 9.3 is retained, because it is a statement about this review and not about the instruments. The earlier versions of this review remain deposited and carry both instruments in full; depositing this version does not withdraw them, and the companion paper states as much rather than undertaking a removal that no deposit can perform.

Three references are added, and their absence was a defect rather than a matter of space. Section 9.1's prior-art paragraph listed TIDieR, the CONSORT extension, SENIAM and Peterchev, and omitted the three consensus minimum reporting standards for individual electrical modalities published in this field since 2021 - RATES for transcranial electrical stimulation (PMID 41044265), REPORT-SCS for spinal cord stimulation in spinal cord injury (PMID 38271712), and the international minimum reporting standards for transcutaneous vagus nerve stimulation (PMID 33854421). Two of the three post-date Section 9 as first drafted; all three were found while preparing the companion paper, and the scope claim made against them is narrower than the one made before. The deposited reference database is extended from 210 to 213 records in the same sitting, and 141 to 144 of its records now carry a reference number for this manuscript.

One thing this note cannot repair from here, said rather than left to be noticed: the summary and novelty paragraphs above were written for version 1.2.6 and are left as they stand. They describe Section 9 as proposing the two instruments, and they name only Peterchev, TIDieR, the CONSORT extension and SENIAM as the related prior work. At version 1.3.0 Section 9 outlines the instruments and cites the companion paper for them, and its prior-art paragraph also names RATES, REPORT-SCS and the transcutaneous vagus nerve stimulation standards.

Two further corrections. A cross-reference was wrong in three places: the four full-text characterisations are listed in Section 11's fifth limitation, not in "Section 9's selection-risk paragraph" - Section 9 has never contained one. And the word count was asserted rather than measured. The header read "approximately 6,500" from version 1.0 onward; the measured figure at version 1.2.6 was 6,826, so the stated number was already 5% low before this version. It now states 7,231, measured, with the counting rule named. Removing two tables does not reduce it, because table rows never counted, while Section 9.1's corrected prior-art paragraph adds prose. No value, table row, figure datum, reference entry, evidence grading or conclusion changes anywhere in this version.

Version history — and why it matters

Version 1.2 (18 September 2026) is a substantial correction of version 1.1, which was graded RED by adversarial internal review and was not fit to circulate. The review found that seven findings unfavourable to surface-applied electrical stimulation had been removed from sources the manuscript otherwise relied on, and that all seven ran in the same direction — toward making the field in which the author has a commercial interest look better founded and safer than its sources say. The manuscript's numbers were accurate: 27 of 27 reference records checked were bibliographically clean. The defect was in selection, not transcription.

All seven are restored in v1.2, each at the point where its source is already cited:

The DepressionDC trial (Burkhardt et al., Lancet 2023) — the largest sham-controlled multicentre trial of transcranial direct-current stimulation in depression, which is null (MADRS difference 0.3, 95% CI −2.4 to 2.9) — had been omitted entirely.

EF-14's mild-to-moderate skin toxicity in 52% of the device group, omitted from a review whose own Table 4 makes skin-effect reporting mandatory.

The pressure-ulcer Cochrane review's reported harms, including limb amputation and death, and its two uncertain outcomes.

ELECT-TDCS's excess of skin redness, tinnitus and nervousness, and its confidence interval.

Faes's analysis of covariance showing the six soft-tissue resistivity means are not statistically distinguishable from one another.

Vöröslakos's conclusion that effective currents are higher than those used in conventional protocols.

Huang's qualifier that 0.8 V/m at 2 mA is the lower limit of effectiveness in animal studies.

A negative trial of tumour-treating fields in recurrent glioblastoma (EF-11) was also added, so that modality is not represented by its positive trial alone. Section 11 of the manuscript records that the omissions occurred, because deleting the evidence of a selection problem would repeat the problem.

Stated limitation on verification

Content verification — checking each quantitative value and each characterisation against its source — was performed at the level of the abstract, and at full text only where the full text was openly accessible. It was not performed against the full text of paywalled sources. Section 1.1 states this, and the manuscript carries an explicit pre-submission condition that the author read those sources in full text before any journal submission. This preprint is deposited with that condition outstanding and openly declared.

Competing interests

The author is the founder, Chief Executive Officer and Chief Technology Officer of Rohera Healthcare & Technology Pvt. Ltd., Pune, India, which develops and markets eMedica, a device combining electrical stimulation with other energy modalities and physiological sensing, and is the named inventor and holder of granted patents relating to it (US 12,226,183 B2; JP 7611901 B2; AU 2020346420 B2; CA 3150348 C; KR 10-2954775 B1; ZA 2022/04073 B), with further grants and pending applications in other jurisdictions.

This review is not neutral ground for its author. Sections 9 and 10 propose reporting requirements and research priorities for a device class that includes the author's own product. No data from that device are presented and no claim of efficacy for it is made anywhere in the review, but readers should weigh the recommendations accordingly. Section 9.3 classifies the review's own principal claims under its own evidence scheme.

Funding

This work received no external funding. It was funded by the author, through Rohera Healthcare & Technology Pvt. Ltd. Self-funding is declared here as a competing interest rather than as the absence of one: no external sponsor, grant reviewer or independent body reviewed, constrained or verified this work at any stage.

Use of AI-assisted tools

A large language model (Claude, Anthropic) was used to run and organise the literature searches, to check bibliographic details against PubMed and Crossref records, to check quoted values against source abstracts and openly accessible full texts, to draft the text, and to conduct the adversarial internal review described above. The author directed this work, reviewed and edited all content, and takes full responsibility for the manuscript. The tool is not an author and does not meet ICMJE authorship criteria.