Acne Scar Procedures During Isotretinoin: What Small Trials Can Establish

By Michael Wiederkehr, M.D.

A person taking isotretinoin may see fewer breakouts while still noticing depressions, uneven texture, or discoloration left by earlier acne. This creates a practical question: should established scars be treated during the course, shortly afterward, or only after a longer waiting period?

The answer cannot be reduced to the word "laser." Fractional ablative treatment, fully ablative resurfacing, nonablative devices, ordinary microneedling, and radiofrequency microneedling expose tissue in different ways. Evidence for one technique cannot automatically authorize another. The exact procedure, patient, protocol, and timing matter.

Two particularly relevant research records address the timing question: a split-face fractional CO2 laser study and a small randomized trial of radiofrequency microneedling combined with low-dose isotretinoin. Their findings are informative, but their size and reporting impose limits. This article explains those limits and places them beside broader reviews, current labeling, and device safety information.

An important tension should be stated at the beginning. Research and expert consensus have questioned a blanket delay for selected procedures, while current isotretinoin product information still advises avoiding resurfacing procedures during treatment and for at least six months afterward. ^[1] A patient should not independently override those instructions after reading a small trial. The evidence belongs in a specialist discussion about the proposed intervention.

Separate active acne, marks, and structural scars

Active inflammatory lesions, residual color changes, and established scars are different treatment targets. A photograph may contain all three, making a single "before and after" impression difficult to interpret. The clinician needs to identify which changes are being treated and how improvement will be measured.

The American Academy of Dermatology distinguishes depressed scars, including ice-pick, boxcar, and rolling patterns, from raised scars. It also distinguishes scars from flat discoloration after acne. ^[2] These differences help explain why one procedure may be inappropriate for another person's apparently similar concern.

The distinction also improves expectations. If a patient mainly wants smoother texture, a reduction in active lesion counts may be welcome but insufficient. If the immediate problem is continuing painful acne, a scar procedure should not obscure the need to control new lesions.

A meaningful treatment plan names the target in ordinary language. "Reduce the depth and visibility of selected depressions" is more precise than "repair all acne damage." Precision helps the patient judge whether the proposed procedure addresses the actual concern.

Why the historical waiting period became controversial

A 2017 systematic review and consensus statement examined 32 publications reporting 1,485 procedures. It found insufficient evidence to support delaying several specified procedures, including fractional ablative and nonablative laser treatments, solely because of current or recent isotretinoin. It did not recommend mechanical dermabrasion or fully ablative laser during systemic treatment. ^[3]

The scope is essential. "Insufficient evidence to support delaying selected procedures" is not equivalent to "all procedures are proved safe." The statement distinguishes techniques, and its evidence base includes different designs rather than a single large randomized safety trial.

A separate ASDS task force reviewed procedures and issued consensus recommendations, also emphasizing distinctions among interventions. ^[4] Such reviews can challenge a broad historical practice while leaving meaningful uncertainty about particular devices, settings, and patients.

A useful analogy is a change from a blanket rule to a more specific question. Removing an automatic exclusion for some procedures does not remove the need to assess healing, pigment changes, scar history, and the exact exposure being proposed.

This is why apparently conflicting advice can coexist. A label may retain a broad precaution; a specialist review may distinguish newer fractional techniques from more extensive resurfacing; a clinic may apply a conservative policy. Patients deserve an explanation of which evidence and instructions apply to the actual plan rather than a declaration that one side is simply outdated.

What "fractional" means for interpreting a study

Fractional treatment addresses portions of the skin in a patterned distribution rather than treating the entire surface in the same way. Ablative and nonablative approaches differ in their interaction with tissue. These technical distinctions are relevant because they define what a trial actually tested.

A patient does not need to choose energy settings or understand every device parameter. However, knowing the procedure's full name prevents an important category error. A study of a fractional CO2 protocol does not establish the safety of fully ablative resurfacing, a deep peel, or another procedure merely advertised as "skin renewal."

Even within a device category, protocols can differ. Coverage, energy, number of passes, session spacing, preparation, and aftercare can influence the exposure. The device name alone is therefore an incomplete description of treatment.

A useful study summary should identify:

• Whether the technique is ablative or nonablative.
• Whether treatment is fractional or full-surface.
• The number and spacing of sessions.
• Whether isotretinoin was concurrent or recently completed.
• The patients included and excluded.
• The outcome scales and follow-up period.

This information does not need to become a do-it-yourself protocol. Its purpose is to judge relevance. If the proposed procedure differs substantially from the studied one, the evidence must be applied cautiously rather than copied as a general permission slip.

The distinction becomes particularly important in marketing, where several technologies may be grouped under one accessible name. Ask for the actual device and treatment category, not just the package title printed on a clinic's website.

The 21-person split-face dataset

The dataset record describes 21 isotretinoin-treated patients in a randomized split-face comparison. One half-face received three monthly fractional ablative CO2 laser sessions. The reported final scar score was 4.71 on the treated side and 9.79 on the comparison side, with lower scores indicating less severe scarring. No unexpected adverse events were reported during the described follow-up. ^[5]

This is promising evidence for the studied combination, not proof that complications cannot occur. The study is small, and a finding of no unexpected events depends on what was monitored, how events were defined, and how long patients were observed.

The split-face design has a practical strength: both sides belong to the same person and share many characteristics. It reduces some between-person variation that complicates a comparison of separate patient groups. Randomizing the side also helps reduce systematic differences in treatment assignment.

However, the two sides are not independent patients. Twenty-one people do not become 42 independent people because two sides were assessed. The analysis should account for the paired design, and a summary should retain the actual participant count.

The public record also represents a particular reporting stage. A dataset description, abstract, and later publication may present different comparisons or follow-up endpoints from related work. They should not be counted automatically as separate independent trials. The next section illustrates why that matters.

The later publication uses a different comparison

The published trial abstract describes concurrent fractional CO2 treatment on one side and the same treatment delayed until six months after isotretinoin cessation on the other. It reports final scar scores of 4.7 versus 7.7, favoring the concurrent side. The authors note standardized settings that were not adjusted for individual skin type as a limitation. ^[6]

Those results should not be silently mixed with the dataset's 4.71-versus-9.79 comparison. The comparison conditions differ: an initially untreated side and a subsequently treated delayed side are not the same endpoint. Reporting both as though they were separate replications would exaggerate the evidence.

This is an original source-comparison issue with direct practical value. When several records share investigators, a treatment protocol, and similar numbers, the reader should check whether they describe stages of the same research. More links do not necessarily mean more independent participants.

The safest summary names the record and comparison beside each number. This makes the evidence traceable and avoids turning an evolving research report into an artificially large body of confirmation.

What the radiofrequency trial reported

The radiofrequency dataset describes a prospective randomized study comparing microneedle radiofrequency plus low-dose isotretinoin with isotretinoin alone. Eighteen participants completed the study, nine per group. The combination group had a reported week-20 lesion clearance rate of 93.02% and a 35-point reduction in ECCA scar score. ^[7]

The first number concerns skin lesions. It does not mean that 93.02% of scars disappeared. The second concerns a scar scoring system. Translating these distinct outcomes into "93% scar removal" would be a substantive error, not a harmless simplification.

The dataset also discusses microbiome observations and a possible mechanism. A change in microbial measurements alongside clinical improvement does not independently prove that the microbial change caused the benefit. Mechanistic hypotheses require evidence beyond the coexistence of two changes.

The underlying deposit contains large image and sequencing materials. Clinical interpretation should distinguish these exploratory biological measurements from the reported patient outcomes. The deposited materials do not by themselves establish that a microbiome change caused a clinical benefit.

With nine completers per group, even a large numerical difference deserves cautious interpretation. Randomization is valuable, but small groups remain vulnerable to imbalance and imprecision. The study can justify further research and a more informed discussion; it cannot establish a universal standard for every device or every isotretinoin regimen.

A current safety communication changes the practical context

FDA issued an October 2025 safety communication about potential serious complications with certain uses of radiofrequency microneedling. Reported concerns include burns, scarring, fat loss, disfigurement, and nerve damage. The communication advises discussing the procedure and provider qualifications and emphasizes that this is a medical procedure. ^[8]

The communication is broader than the isotretinoin trial and does not establish an isotretinoin-specific interaction. It nevertheless matters when someone uses a small favorable trial to claim the device category has no serious risks. The trial and the safety notice address different evidence and must both remain visible.

A useful interpretation separates three questions:

1. Can the procedure improve the studied outcome?
2. What complications can occur with the procedure in practice?
3. Does current or recent isotretinoin alter the balance for this patient and protocol?

A favorable answer to the first does not settle the other two. Likewise, awareness of serious reported complications does not provide their incidence for a particular protocol without suitable data.

Patients should not assume that "minimally invasive" means risk-free or appropriate for home use. A device that penetrates tissue or delivers energy needs appropriate assessment, training, and follow-up. This is relevant to the procedure itself, independently of whether the patient takes isotretinoin.

The appropriate response is a specific consultation about the proposed intervention, rather than a generalized endorsement or rejection of every device sharing a marketing term.

Newer observational evidence still has uncertainty

A 2026 retrospective cohort studied fractional CO2 resurfacing in 106 patients with Fitzpatrick skin types III-VI. Delayed re-epithelialization beyond 14 days occurred in 11 patients, with a numerically higher proportion in the concurrent isotretinoin group. The adjusted estimate was not statistically significant and had a wide confidence interval. ^[9]

This finding illustrates why "not statistically significant" is not a synonym for "proved identical safety." A wide interval may remain compatible with clinically meaningful differences. The appropriate reading preserves both the observed pattern and the uncertainty.

The retrospective design also differs from a randomized trial. Clinicians selected treatment and timing under real-world circumstances. The groups may differ in ways that were incompletely measured or adjusted. That limits causal interpretation even when a statistical model is used.

The study adds information about a population important to pigment and healing discussions. It should not be used to declare that skin type no longer matters or that a single protocol applies to all patients. Inclusion of several skin types is useful, but subgroup numbers and the actual observed outcomes still matter.

Taken together, the newer research supports increasingly specific questions. It does not support replacing an overly broad historical rule with an equally broad claim of universal safety. Progress in evidence often means better differentiation, not the elimination of every precaution.

Why small trials cannot exclude uncommon complications

A study with no observed event may still be too small to detect an uncommon risk. The probability of observing an event depends on the actual risk, sample size, independence of observations, and follow-up. "We saw none" is a report of observations, not proof that the risk is zero.

An invented calculation makes the issue concrete. If a complication truly occurred in one of every 100 independently treated people, a 21-person study could easily observe no cases. Under a simplified constant-risk model, the chance of zero events would be approximately 0.99 raised to the 21st power, or about 81%. These are hypothetical figures, not estimated rates for fractional laser treatment.

The example does not mean a study must include thousands before it has value. Small studies can identify feasibility, early efficacy, expected reactions, and problems that occur relatively often. They can also refine the design of larger studies.

What they cannot do is make an absence-of-events claim stronger than their size allows. This is especially important when a complication is serious but uncommon. A reader should distinguish "no event observed in this sample" from "the procedure cannot cause the event."

The same principle applies to follow-up. A short observation window cannot establish the absence of a later outcome. A longer period helps, but only if enough participants remain under observation and the relevant events are actively sought.

Compare outcomes before comparing procedures

The fractional laser and radiofrequency studies used different designs, scoring systems, protocols, and populations. Their percentages cannot be placed in a simple league table. A larger percentage does not necessarily mean a better procedure when the underlying outcomes differ.

The last row deserves attention. If a patient receives medication plus a procedure, the combined result does not reveal what each component would have achieved alone unless the design supports that comparison. A control group can help, but the exact contrast remains essential.

A meaningful comparison between procedures would ideally use similar patients, relevant outcomes, suitable follow-up, and direct allocation. Without that, cross-study comparisons should be described as exploratory rather than as proof that one technology is superior.

For patients, the practical comparison is even more individual: which option addresses the actual scar pattern, fits the person's risk profile, and offers a reasonable balance of benefit, cost, and recovery burden?

Scar type should guide the treatment discussion

The AAD explains that different scar patterns may need different approaches and that treatment plans consider features such as scar type, number, skin tone, age, and budget. ^[10] This helps explain why a procedure that benefits a study average may not be the best first option for every individual scar.

The examination should identify whether a patient has a uniform pattern or several patterns together. A mixed pattern may lead to a staged or combined plan rather than repeating one device over the entire face without a specific rationale.

Patients can ask the clinician to identify the scars that are likely to respond and those that may respond less well. This creates a more realistic expectation than a general percentage offered before examination. It also helps explain why the plan may change after the first stage is assessed.

The language should remain concrete. "This treatment is intended to reduce the visibility of these depressions" gives the patient a target. "Collagen remodeling" may describe a proposed mechanism, but it does not by itself tell the patient which concern will improve or how much.

A plan can also distinguish structural change from color improvement. If photographs look better because redness has faded, that benefit should be acknowledged without claiming an equivalent change in scar depth. Honest description supports both realistic satisfaction and valid comparison over time.

Skin tone, pigmentation, and healing history

Pigment changes are among the possible adverse effects of microneedling and other procedures, and FDA information advises considering relevant skin characteristics and risk factors. ^[11] A history of unusual scars or previous procedural reactions also belongs in the consultation.

The useful question is not whether a device is advertised as suitable for "all skin types." It is what evidence, settings, preparation, and aftercare support its use for this particular patient. A broad promotional statement cannot replace an individual assessment.

Patients should mention previous hyperpigmentation, hypopigmentation, raised scars, delayed healing, infections, and reactions to earlier treatments. The clinician needs the actual history rather than a self-assigned label of "sensitive skin" alone.

Photographs of previous reactions can help if the original medical records are unavailable, but they do not establish the diagnosis by themselves. The timeline, treatment details, and any assessment at the time remain useful context.

This section is not an argument that people with a particular skin tone should be excluded automatically. It is an argument for recognizing differences that may affect the plan and uncertainty. Underrepresentation in a trial should be discussed as an evidence limit rather than hidden behind a claim that the device works identically for everyone.

A good consultation explains both the anticipated benefit and the relevant complications in terms the patient can understand. The opportunity to ask questions should come before a package is purchased or treatment begins.

Reconciling current labeling with procedural research

The current isotretinoin product information advises avoiding wax epilation and resurfacing procedures during treatment and for at least six months afterward because of possible scarring. ^[1] The research discussed above does not erase that wording.

A specialist may consider newer evidence for a precisely defined procedure, but that discussion should acknowledge the label rather than imply it no longer exists. The patient should understand why a proposed plan differs from broad instructions, what evidence supports it, and what uncertainty remains.

This is a shared clinical decision, not a loophole for a patient to exploit. A clinic's willingness to offer an earlier procedure does not alone establish its appropriateness. Conversely, a clinic that follows a conservative waiting policy should be able to explain that policy without pretending no newer evidence exists.

Several practical questions help:

• Does the proposed procedure fall within the label's broad precaution?
• What research applies to the exact technique and timing?
• How similar were the trial patients to this patient?
• Is there a reason to proceed now rather than later?
• What are the alternatives, including waiting?
• How will complications and aftercare be managed?

The answers may support postponement or a specialist-supervised plan. A general educational article should not decide that balance remotely. Its role is to make the discussion more informed and the claims more accurate.

Timing is more than a calendar interval

The timing question includes current acne activity, skin condition, treatment stage, recovery needs, and the patient's goals. A date alone does not establish readiness. Two people six months after the last capsule may have different skin findings, scar patterns, and medical histories.

Likewise, a study that starts a procedure during a particular regimen does not show that every point in every regimen is equivalent. Daily dose, duration, co-treatments, and patient selection can all affect relevance even when the article title uses the same medicine name.

A useful timing plan identifies why now is appropriate or why waiting is preferable. It should not be based solely on an upcoming event, a discounted package, or an assumption that earlier treatment must always produce a better outcome.

Patients should also consider the practical recovery period. Even a procedure described as having limited downtime may affect work, social plans, sun exposure, or the ability to follow aftercare. These are legitimate considerations in the benefit-risk discussion rather than trivial inconveniences.

If treatment is postponed, the patient should understand the next review point and what will be reassessed. "Wait indefinitely" is less useful than a clear explanation of what needs to settle, what treatment continues meanwhile, and when the scar plan can be reconsidered.

The goal is not to find the earliest technically possible date. It is to choose a defensible time for a procedure that addresses the actual problem under a workable care plan.

Preparing for a consultation

Bring the isotretinoin product name, current prescribed dose, start date, any planned end date, and other medicines or skin products. Describe current dryness, irritation, wounds, or previous treatment reactions. The procedural clinician should know about the course even if another service prescribes it.

A useful consultation should include a scar assessment rather than moving directly from a request for a named device to booking. Patients often arrive after seeing a procedure online, but the desired outcome should guide the choice.

If the answers rely entirely on before-and-after photographs or a device manufacturer's brochure, ask for the relevant clinical evidence and the clinician's own explanation. The point is not to demand certainty, but to ensure that uncertainty is acknowledged before treatment rather than after a disappointing result.

Before-and-after photographs need context

Photography can document a visible concern, but lighting, angle, expression, distance, focus, and skin preparation can alter how scars appear. A photograph is more informative when conditions are standardized and the timing is clear.

This matters particularly for texture. Side lighting can emphasize depressions, while diffuse frontal lighting can make them less visible. A change between photographs can therefore reflect both treatment and imaging conditions. The reader should not assume that every visible difference is biological improvement.

Ask whether photographs were taken before and after a full recovery interval, whether makeup or filters were used, and whether active acne and color changed alongside scar texture. These questions do not accuse the provider of deception. They establish whether the comparison answers the patient's actual question.

A clinic's best result is also not the expected result for every patient. A useful consultation explains the range of outcomes and the relevance of the displayed case. Similar age or skin tone alone does not guarantee a similar scar pattern, treatment protocol, or response.

Patients should avoid judging the final result too early or under constantly changing conditions. The clinician should explain when assessment is meaningful and how photographs will be used. A consistent record can support discussion, but it should not encourage daily scrutiny that cannot distinguish ordinary fluctuation from a treatment effect.

Aftercare and the response to an unexpected course

AAD patient information emphasizes that laser treatment involves preparation, aftercare, and realistic expectations rather than a procedure alone. It also notes that laser treatment does not simply erase a scar. ^[12] These principles remain relevant when the timing overlaps with medication treatment.

Before the procedure, the patient should know which products to use, which to avoid, how to protect the treated area, and what appearance is expected during recovery. The plan must come from the treating clinician and device-specific procedure, not a generalized routine borrowed from another treatment.

The patient should also know what development requires contact and which service will respond. If the recovery differs from the expected course, a photograph may help communication, but it should not replace assessment when the clinician recommends one.

A useful aftercare plan identifies a review date and a route for earlier concerns. It should not require the patient to decide whether a complication is "serious enough" without any guidance. Clear instructions reduce both delayed reporting and unnecessary uncertainty.

If another clinician is managing isotretinoin, the medication plan should be coordinated. The patient should not receive contradictory instructions from two services without a route to resolve them. A short written record of the agreed plan can be valuable when care is shared.

The procedure is therefore a care pathway rather than a single appointment. Its quality includes patient selection, technique, aftercare, access, and follow-up, not only the device used on the day.

Interpreting cost and value without a device popularity contest

The cheapest session is not necessarily the least expensive course, and the most expensive technology is not necessarily the most appropriate. A comparison should consider the target, expected number of sessions, recovery, follow-up, and what happens if the response is incomplete.

A hypothetical package costing less per session may require more sessions or exclude follow-up. Another may include review but still be inappropriate for the scar pattern. Price becomes meaningful only after the treatment being purchased is clearly defined.

Ask whether the estimate includes consultation, anesthesia where relevant, aftercare products, planned reviews, and management of complications. Do not assume that an initial quote represents the full cost or that additional sessions will definitely be needed. The clinician should explain which parts are predictable and which depend on response.

The outcome should also be defined before discussing value. If the goal is complete removal of every scar, even a technically successful procedure may disappoint because the expectation was unrealistic. If the goal is a specific degree of visible improvement, the plan can be evaluated more honestly.

This is an economic decision as well as a clinical one, but financial pressure should not determine safety. A time-limited discount does not make an uncertain procedure more appropriate during a medication course. Patients should have enough time to understand the plan without being pushed into an immediate commitment.

Frequently asked questions

What a defensible treatment proposal should contain

A defensible proposal names the problem, the intervention, and the reason for its timing. It acknowledges the current isotretinoin precaution and explains how the cited evidence applies. It also states what remains uncertain rather than relying on the word "safe" without qualification.

The benefit estimate should concern the patient's actual target. If a clinic quotes a lesion-clearance percentage to advertise scar removal, the patient has not received a meaningful estimate. If it quotes a scar score, the clinician should explain what that score means in visible terms and why it is relevant.

The safety discussion should distinguish expected reactions, uncommon complications, and issues specific to the patient's history. It should not claim that a small sample excludes rare events. Nor should it use a device safety communication to invent an incidence that the reporting data cannot establish.

The practical plan should include preparation, recovery, review, and access if the course differs from expectations. Costs should describe the actual package. Photographs should support a consistent assessment rather than serve as the entire evidence base.

Finally, the patient should understand that waiting is also an option to discuss. A decision to delay can be informed and purposeful. A decision to consider earlier treatment should be equally explicit about its rationale and uncertainty. The quality of the decision is measured by its relevance to the patient and evidence, not by how quickly a procedure can be booked.

A worked example of reading a procedural claim

Consider a fictional advertisement: "A clinical study showed 93% clearance, so our resurfacing package removes almost all acne scars safely during Accutane." Several separate claims have been joined together. The first task is to identify the study, its actual endpoint, and the device being offered. If the number describes lesion clearance, it cannot be used as a scar-removal estimate. If the study used radiofrequency microneedling, it cannot automatically validate an unrelated resurfacing device.

The next task is to inspect the comparison. A combination regimen compared with isotretinoin alone addresses an added-treatment question. It does not establish that every component of the package is necessary or that the package is superior to every alternative. A small group of completers also does not support a universal claim about rare complications.

The word "safely" needs its own examination. Does the claim mean that no unexpected events were observed in the study, that a regulatory authority reviewed the exact use, or that the clinic promises no complications? These are not interchangeable statements. A responsible provider should explain the actual meaning without relying on the reassuring sound of the word.

Then compare timing and patient selection. A protocol involving a particular low-dose course and selected participants is not direct evidence for every dose, age, skin condition, or healing history. If those differences are relevant, the clinician should describe how they affect the decision.

Finally, bring current labeling back into the discussion. The existence of a favorable study does not remove the instruction in the patient's product information. A defensible specialist proposal acknowledges that issue and explains its rationale rather than asking the patient to ignore it.

The corrected version of the fictional claim would be much narrower: "A small trial reported improved lesion and scar outcomes with a specified combination. Its findings can inform a specialist assessment, but they do not predict complete scar removal or exclude uncommon complications." That wording remains informative. It simply stops before making claims the study cannot support.

This example is useful beyond one device. Whenever a percentage appears in procedural marketing, ask which outcome it measures, which patients produced it, what the comparison was, and whether the offered intervention matches the tested one. Those questions turn a striking number into evidence that can actually be evaluated.

A final check concerns the source date and version. A dataset can be updated after its first release, and a publication can describe later follow-up. Record which version supports the claim. If the underlying participants overlap, describe the reports as related evidence rather than adding their sample sizes. This simple source discipline prevents an article from appearing more certain merely because the same investigation is available through several links.

Matching the procedure to the baseline concern

Before a procedure, ask the clinician to document the scar pattern and the particular changes the treatment is intended to improve. This creates a useful baseline for the later review. If the main concern is a small number of deeper depressions, a change in overall facial redness may be welcome without answering whether those depressions improved.

The baseline discussion should also distinguish the treated area from the whole face. A study or photograph may assess a selected region, while a patient expects an equal effect everywhere. The clinician can explain which areas are included, which may need another approach, and how progress will be judged after recovery. This keeps the expected outcome tied to the actual procedure rather than a broad promise about skin renewal.

References

1. Isotretinoin capsules prescribing information, revised August 2026. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=1cf11710-f966-4529-8e08-02175f588bca
2. American Academy of Dermatology. Acne scars: Signs and symptoms. https://www.aad.org/public/diseases/acne/derm-treat/scars/symptoms/
3. Spring LK, et al. Isotretinoin and Timing of Procedural Interventions: A Systematic Review With Consensus Recommendations. JAMA Dermatology. 2017. https://pubmed.ncbi.nlm.nih.gov/28658462/
4. ASDS Guidelines Task Force. Consensus Recommendations Regarding the Safety of Lasers, Dermabrasion, Chemical Peels, Energy Devices, and Skin Surgery During and After Isotretinoin Use. 2017. https://pubmed.ncbi.nlm.nih.gov/28498204/
5. Fractional ablative CO2 laser treatment for acne scars in patients receiving isotretinoin: dataset description. https://data.mendeley.com/datasets/dkbghtbwgs/1
6. Taleb E, et al. Fractional ablative CO2 laser and oral isotretinoin: a prospective randomized controlled split-face trial comparing concurrent versus delayed laser treatment for acne scars. Lasers in Surgery and Medicine. 2024. https://pubmed.ncbi.nlm.nih.gov/37555247/
7. A Prospective Randomized Trial of Microneedle Radiofrequency Combined with Low-dose Isotretinoin vs. Isotretinoin Alone for Acne Vulgaris Treatment. Dataset version 3. https://data.mendeley.com/datasets/rnkczyppy4/3
8. FDA. Potential Risks with Certain Uses of Radiofrequency Microneedling. October 15, 2025. https://www.fda.gov/medical-devices/safety-communications/potential-risks-certain-uses-radiofrequency-rf-microneedling-fda-safety-communication
9. Ablative Fractional CO2 Laser Resurfacing for Atrophic Acne Scars During and After Oral Isotretinoin: A Single-Center Retrospective Cohort Study in Fitzpatrick III-VI Patients. 2026. https://pubmed.ncbi.nlm.nih.gov/42430738/
10. American Academy of Dermatology. Acne scars: Consultation and treatment. https://www.aad.org/public/diseases/acne/derm-treat/scars/treatment
11. FDA. Microneedling Devices. https://www.fda.gov/medical-devices/aesthetic-cosmetic-devices/microneedling-devices
12. American Academy of Dermatology. 10 things to know before having laser treatment for your scar. https://www.aad.org/public/cosmetic/scars-stretch-marks/laser-treatment-scar