This cohort study assesses patient impressions, outcomes, and recovery time after clinic-based surgery for hidradenitis suppurativa and the association of patient characteristics with outcomes.
Key Points
Question
What are the outcomes, patient impressions, and recovery period after clinic-based procedures for hidradenitis suppurativa (HS)?
Findings
This cohort study of 194 surgical procedures among 78 patients who underwent clinic-based HS surgery found a self-reported recurrence rate of 41%, but patient satisfaction was high, pain was tolerable, and recovery periods were relatively brief.
Meaning
In this study, clinic-based surgery for HS was well tolerated, was associated with high rates of patient satisfaction, and allowed for return to work and normal activity within days for most patients.
Abstract
Importance
Surgical intervention is frequently needed to treat hidradenitis suppurativa (HS). Patient satisfaction is high based on previous studies, but reports of patient impressions of clinic-based operative experiences and postoperative recovery are limited.
Objective
To characterize patient impressions, outcomes, and recovery time after clinic-based surgical treatment of HS and examine patient characteristics associated with outcomes.
Design, Setting, and Participants
This retrospective cohort study included patients 12 years or older who underwent clinic-based surgical procedures for treatment of HS at a single subspecialty HS clinic at the University of North Carolina Department of Dermatology from April 2014 to December 2018. Data analysis was performed from January to September 2021.
Exposures
Clinic-based deroofing and excisional procedures performed as part of routine care.
Main Outcomes and Measures
The primary outcomes were patient-reported recurrence of HS at the site of surgery, patient satisfaction with the procedures and outcomes, and patient-reported pain and recovery associated with surgery obtained from electronic medical record review and patient questionnaires.
Results
Outcomes of 194 procedures for 78 patients (65 [83%] female; mean [SD] age, 35.1 [12.1] years) were analyzed. Self-reported rate of recurrence was 41% (79 procedures). Despite recurrence, most patients (148 procedures [76%]) were very satisfied with their surgical results. The median number of missed days of work or school was 2 (IQR, 1-7 days), and the median number of days until return to normal activity was 10 (IQR, 3-14 days). In addition, for 126 of the 194 procedures (65%), patients stated that pain during an HS flare was worse than pain during surgical recovery.
Conclusions and Relevance
In this cohort study, patients reported high rates of satisfaction with clinic-based HS surgery. Recovery was typically rapid, with most patients rating postsurgical pain as less severe than their HS pain.
Introduction
Hidradenitis suppurativa (HS) is a chronic inflammatory skin condition consisting of painful abscesses, nodules, sinus tracts, and scarring with a predilection for intertriginous sites.1,2 Hidradenitis suppurativa is more common in Black individuals, female individuals, and people with low socioeconomic status; onset of disease occurs most often in the third and fourth decades of life.3,4,5,6,7 Typical medical treatment includes antibiotics, hormonal therapies, and tumor necrosis factor inhibitors, but surgery is often vital in treating recurrent nodules and tunnels.8,9,10,11,12,13,14,15
Deroofing involves removal of overlying skin at the sites of sinuses, and the mostly dermal wound base is left to heal by secondary intention.16 Excision removes affected skin to the subcutaneous fat with a small margin of healthy tissue.17,18 The excised area may be left to heal by secondary intention, closed primarily, or covered by a flap or graft. Rates of reported recurrence have been similar for deroofing and excision, with rates ranging from less than 5% to 25%.19,20,21,22,23 Few data comparing primary closure vs secondary intention healing have been reported.24 Incision and drainage relieves acute pain from abscesses, but excision or deroofing is the definitive treatment for longstanding or recurrent sinuses.16,19,20,25
Patient satisfaction and outcomes after HS surgery performed with general or local anesthesia have been positive in retrospective questionnaire-based studies, including studies examining excision, deroofing, and a combination of approaches.16,21,22,23,26,27,28 The results from these studies indicated that patients were very satisfied with the scar appearance and would choose and recommend surgery as a treatment modality for HS. Most studies evaluating patient satisfaction16,21,22,23,26,27,28 have been conducted in populations in which most patients were White or have included no mention of race or ethnicity, but it is well documented that HS is more prevalent in Black patients.29,30 In addition, few studies have explored the association of postoperative recovery with missed workdays and time to wound healing.
In this cohort study, we aimed to analyze patient outcomes and satisfaction after HS surgery and to assess the association of demographic information, lifestyle factors, and treatment choices with surgical outcomes. We also describe patient impressions of procedures and recovery, which have not been widely explored previously to our knowledge.
Methods
Study Design, Data Collection, and Variables
In this retrospective cohort study, the population included patients 12 years or older who met clinical criteria for HS and underwent surgical intervention at the University of North Carolina (UNC) Department of Dermatology in Chapel Hill. This study was reviewed and approved by The University of North Carolina at Chapel Hill institutional review board. Written informed consent was obtained when interviews were conducted in person; oral consent was obtained during phone interviews. Data were deidentified. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
We used clinical tracking lists and the UNC i2b2 database to identify patients with a diagnosis of HS who had deroofing or excision performed with the use of local anesthesia in a clinic setting between April 2014 and December 2018. We used International Classification of Diseases, Ninth Revision (705.83); International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (L73.2); and Current Procedural Terminology codes (11450, 11451, 11462, 11463, 11470, and 11471) for HS surgery. This retrospective cohort study involved combining a query of the Carolina Data Warehouse with survey methods to analyze patient experience and rates of HS recurrence after surgical treatments. Pencil-and-paper questionnaires were created and administered in the clinic or by telephone. To reduce the risk of biasing patients toward positive responses, questionnaires were administered by trained research assistants, and respondents’ identities were concealed from the principal investigator (C.S.) who had performed the procedures. Clinical data were mined from the UNC electronic medical records system. Data analysis was performed from January 2021 to September 2021.
The primary interventions were surgical deroofing or local excision performed with the use of local anesthesia in an outpatient dermatology clinic. Deroofing was performed by probing sinuses in the affected areas, opening them with scissors or a scalpel, and removing the tissue overlying the sinuses. Sequential probing and deroofing of any tunnels at the margin of a wound were performed until the area was fully treated. For excisions, probing to delineate involved areas was performed, and tissue was excised with a 5- to 10-mm margin to the subcutaneous fat and typically repaired with a layered closure.
Each patient was allowed to contribute more than 1 surgical experience to the sample, which was accounted for by using a mixed-effects analysis when possible. The primary outcomes were patient-reported recurrence of HS at the site of surgery, patient satisfaction with the procedures and outcomes, and patient-reported pain and recovery associated with surgery. Satisfaction and pain scores were self-reported on a Likert scale of 1 to 10. Reported scores greater than 5 were considered as the binary outcome of “satisfaction” for logistic regression analyses. Recurrence was assessed by asking whether patients had “new areas of HS arise within the surgical scar” after the procedure. Surgical sites were categorized as axilla; inframammary; inguinal and pubic; labia, scrotum, and perianal; buttock; thigh; neck; and other. Hurley stage and use of tumor necrosis factor inhibitors and antibiotics at the time of surgery were used to represent an advanced HS disease state. For regression modeling, strength referred to the absolute value of the β coefficient in a bivariate logistic regression, and the highest of the 3 β coefficients was considered to be the strongest and was used in multivariable analyses.
Statistical Analysis
Logistic regression models estimated bivariate and multivariable-adjusted odds ratios (ORs) and 95% CIs of patient-reported recurrence associated with patient demographic factors, clinical risk factors, HS severity, and surgery characteristics. Demographics were retrieved from the medical records, and patients could choose not to report certain items. A final model was reduced to the minimum number of variables that kept the overall model within statistical similarity to the fully adjusted model based on nonsignificant likelihood ratio tests.
Mixed-effects multiple linear regression models estimated β coefficients and 95% CIs to assess the association between patient-reported recovery variables, including days of missed work or school after surgery, days to return to normal activities, and days for skin to completely heal, and patient demographic factors, clinical risk factors, HS severity, and surgery characteristics. Because many patients underwent multiple surgical procedures, in the mixed-effects model, the individual patient was considered a random effect to reduce bias from patients who underwent multiple procedures. Days of missed work or school and days to return to normal activity were natural log transformed to achieve a normal distribution for linear regression. Similarly, days for skin to completely heal required square root transformation. For patient-reported days to return to work, days to return to normal activity, and days required for complete healing, values greater than 45, 60, and 90, respectively, were considered erroneous responses and treated as missing rather than outliers. We believed that because of the phrasing of the question, these respondents answered with regard to their general HS disease rather than the surgery itself.
Logistic regression models were used to estimate bivariate ORs and 95% CIs of patient satisfaction associated with patient demographic factors, clinical risk factors, HS severity, and surgery characteristics. Because of frequency limitations of patients not satisfied with surgical procedures or outcomes, it was unfeasible to conduct multivariable-adjusted or mixed-effects analyses.
In cases of nonconvergence of mixed-effects modeling, a sensitivity analysis was performed repeating the analyses after limiting the data to only the most recent procedure for each patient. Although we performed multiple comparisons simultaneously, the Bonferroni correction was not applied because in this exploratory analysis, the goal was to identify important variables of clinical significance and reduce the risk of type II error in which the null hypothesis—in this case, lack of an association with a variable—was accepted even though it was false.31,32,33,34
All statistical analyses were performed using Stata/MP, version 16.0 (StataCorp LLC), with a 2-sided P < .05 considered to be statistically significant.
Results
Of the 201 identified patients who met clinical criteria and underwent surgical intervention, 78 (39%) completed surveys (65 [83%] female; mean [SD] age, 35.1 [12.1] years) and were included in the analysis of outcomes; data on 194 independent surgical procedures were collected for these respondents. Table 1 summarizes patient demographic and clinical characteristics, surgical sites, and procedure types stratified by recurrence rates and patient satisfaction. Of the 194 procedures, 148 (76%) were performed among female patients. Most procedures were performed among non-White patients (Black, American Indian or Alaska Native, Asian [including Indian], and Native Hawaiian or Other Pacific Islander; 106 [55%]) and patients who had obesity class 1 (67 [35%]), did not currently smoke (163 [84%]), and did not have diabetes (128 [66%]). In addition, most procedures were performed among patients with more advanced disease, indicated by 117 (60%) with Hurley stage 3 disease, 101 (52%) receiving a tumor necrosis factor inhibitor at the time of surgery, and 102 (53%) receiving antibiotics at the time of surgery. The most common procedure was deroofing (129 procedures [66%]), and the axilla was the most common site (64 procedures [33%]).
Table 1. Patient, Clinical, and Surgery Characteristics by Patient-Reported Recurrence and Satisfaction.
| Characteristic | Surgical procedures, No. (%) | |||
|---|---|---|---|---|
| Total | Recurrence | Satisfaction with procedure | Satisfaction with outcome | |
| No. (%) | 194 (100) | 79 (41) | 169 (87) | 166 (85) |
| Age at time of surgery, y | ||||
| <25 | 47 (24) | 26 (55) | 37 (79) | 34 (72) |
| 25-35 | 70 (36) | 36 (66) | 59 (84) | 58 (83) |
| >35 | 77 (40) | 17 (22) | 73 (95) | 74 (96) |
| Sex | ||||
| Female | 148 (76) | 61 (41) | 123 (83) | 122 (82) |
| Male | 46 (24) | 18 (39) | 46 (100) | 44 (100) |
| Race and ethnicity | ||||
| Non-Whitea | 106 (55) | 46 (43) | 90 (85) | 85 (80) |
| White | 65 (34) | 29 (45) | 56 (86) | 59 (91) |
| Missingb | 23 (12) | 4 (17) | 23 (100) | 22 (96) |
| BMIc | ||||
| Normal | 29 (15) | 16 (55) | 21 (72) | 23 (79) |
| Overweight | 14 (7) | 10 (71) | 9 (64) | 11 (79) |
| Obesity class 1 | 67 (35) | 22 (33) | 67 (100) | 64 (96) |
| Obesity class 2 | 23 (12) | 8 (35) | 15 (65) | 18 (78) |
| Obesity class 3 | 51 (26) | 22 (43) | 47 (92) | 40 (78) |
| Missingd | 10 (5) | 1 (10) | 10 (100) | 10 (100) |
| Current smoker | 31 (16) | 15 (48) | 19 (61) | 25 (81) |
| Diabetes | 66 (34) | 25 (38) | 57 (86) | 59 (89) |
| Hurley stage | ||||
| 2 | 77 (40) | 32 (42) | 66 (86) | 68 (88) |
| 3 | 117 (60) | 47 (40) | 103 (88) | 98 (84) |
| TNF inhibitor | 101 (52) | 49 (49) | 89 (88) | 82 (81) |
| Antibiotic therapy | 102 (53) | 45 (44) | 90 (88) | 93 (91) |
| Surgery type | ||||
| Deroofing | 129 (66) | 43 (33) | 112 (87) | 111 (86) |
| Local excision with closure | 53 (27) | 27 (51) | 50 (94) | 45 (85) |
| Local excision without closure | 12 (6) | 9 (75) | 7 (58) | 10 (83) |
| Surgical location | ||||
| Axilla | 64 (33) | 23 (36) | 59 (92) | 52 (81) |
| Inframammary | 24 (12) | 13 (54) | 21 | 18 (11) |
| Inguinal or pubic region | 47 (24) | 20 (43) | 37 (79) | 42 (89) |
| Labia, scrotum, or perianal region or perineum | 19 (10) | 3 (16) | 18 (95) | 19 (100) |
| Buttock, thigh, neck, or other | 40 (21) | 20 (50) | 34 (85) | 35 (88) |
| Time from surgery to questionnaire, mo | 13 (6-25) | 16 (7-26) | 13 (7-25) | 13 (7-24) |
| Recurrence | 79 (41) | NA | 63 (80) | 62 (78) |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); NA, not applicable; TNF, tumor necrosis factor.
The non-White category included the following racial and ethnic groups: Black, American Indian or Alaska Native, Asian (including Indian), and Native Hawaiian or Other Pacific Islander.
Race and ethnicity data were missing for 7 patients.
Normal BMI was considered 20.0 to 24.9; overweight, 25.0 to 29.9; obesity class 1, 30.0 to 34.9; obesity class 2, 35.0 to 39.9; and obesity class 3, 40 or greater.
Body mass index was missing for 3 patients.
Recurrence
In the total cohort, the rate of patient-reported recurrence of disease at the previous surgical site was 41% (79 procedures). Table 2 presents bivariate and multivariable-adjusted ORs of patient-reported recurrence associated with patient and surgical characteristics. In bivariate analyses, there was a negative association between recurrence and age greater than 35 years (OR, 0.21; 95% CI, 0.10-0.48; P = .01). The final reduced logistic regression model showed age and surgery type to be significantly associated with recurrence. Patients older than 35 years were significantly less likely to experience recurrence (OR, 0.19; 95% CI, 0.08-0.49; P = .003) compared with those younger than 25 years. Odds of recurrence were 4.38 (95% CI, 1.02-18.29; P = .04) times higher among patients who had excisions without closure compared with those who underwent deroofing. There was no statistically significant difference in odds of recurrence between procedures with excisions with vs without closure.
Table 2. Logistic Regression of Patient-Reported Recurrence Associated With Patient Demographics, Clinical Risk Factors, and Surgery Characteristics.
| Characteristic | Odds ratio (95% CI) | ||
|---|---|---|---|
| Bivariate | Fully adjusted modela | Reduced modelb | |
| Age at time of surgery, y | |||
| <25 | 1 [Reference] | 1 [Reference] | 1 [Reference] |
| 25-35 | 0.88 (0.41-1.90) | 0.98 (0.32-2.85) | 0.62 (0.26-1.45) |
| >35 | 0.21 (0.10-0.48)c | 0.24 (0.08-0.74)c | 0.19 (0.08-0.49)c |
| Sex | |||
| Female | 1 [Reference] | [Reference] | NA |
| Male | 0.84 (0.42-1.65) | 1.23 (0.43-3.51) | NA |
| Race and ethnicity | |||
| Non-Whited | 1.02 (0.55-1.91) | 0.68 (0.28-1.62) | NA |
| White | 1 [Reference] | [Reference] | NA |
| BMIe | |||
| Normal | 1 [Reference] | 1 [Reference] | NA |
| Overweight | 2.03 (0.52-8.00) | 1.79 (0.32-10.1) | NA |
| Obesity class 1 | 0.40 (0.16-0.97) | 0.28 (0.07-1.06) | NA |
| Obesity class 2 | 0.43 (0.14-1.34) | 0.34 (0.07-1.68) | NA |
| Obesity class 3 | 0.62 (0.25-1.54) | 0.43 (0.12-1.61) | NA |
| Current smoker | 1.34 (0.62-2.91) | 0.84 (0.24-2.93) | NA |
| Diabetes | 0.81 (0.44-1.50) | 1.57 (0.66-3.74) | NA |
| TNF inhibitor | 1.79 (0.99-3.25) | 1.93 (0.87-4.33) | NA |
| Surgery type | |||
| Deroofing | 1 [Reference] | 1 [Reference] | 1 [Reference] |
| Local excision with closure | 2.01 (1.04-3.88) | 1.81 (0.72-4.52) | 1.49 (0.71-3.15) |
| Local excision without closure | 5.57 (1.43-21.7) | 4.56 (0.82-25.2) | 4.38 (1.02-18.29)c |
| Surgical location | |||
| Axilla | 1 [Reference] | 1 [Reference] | NA |
| Inframammary | 2.25 (0.84-5.97) | 3.48 (1.04-11.7)c | NA |
| Inguinal or pubic region | 1.57 (0.70-3.50) | 2.24 (0.74-6.80) | NA |
| Labia, scrotum, or perianal region or perineum | 0.32 (0.08-1.24) | 0.47 (0.09-2.45) | NA |
| Buttock, thigh, neck, or other | 1.73 (0.77-3.89) | 2.25 (0.76-6.61) | NA |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); NA, not applicable; TNF, tumor necrosis factor.
Adjusted for age, sex, race and ethnicity, BMI, smoking status, diabetes, TNF inhibitor use, surgery type, and surgical location.
Adjusted for age and surgery type. The model was reduced by removing variables that jointly did not significantly alter the logistic regression model (likelihood ratio test, P = .09; n = 167 owing to missing variables for BMI and race and ethnicity).
Significant at P < .05.
The non-White category included the following racial and ethnic groups: Black, American Indian or Alaska Native, Asian (including Indian), and Native Hawaiian or Other Pacific Islander.
Normal BMI was considered 20.0 to 24.9; overweight, 25.0 to 29.9; obesity class 1, 30.0 to 34.9; obesity class 2, 35.0 to 39.9; and obesity class 3, 40 or greater.
There was no statistically significant difference in odds of patient-reported recurrence associated with sex, race and ethnicity, body mass index, diabetes status, tumor necrosis factor inhibitor use, or surgical site. Although surgeons are often concerned about the effect of smoking on wound healing and outcomes,35 there was not a significant association between smoking and recurrence rates.
Owing to nonconvergence of statistical modeling with a mixed-effects logistic regression using patients as a random effect, it remained unclear whether removing potential bias from multiple reports per patient changed the significant associations. A sensitivity analysis limiting the sample to only the most recent surgery per patient yielded similar estimations; however, likely because of the smaller sample size, findings were no longer significant.
Recovery
Patients reported 3 recovery metrics: number of days to return to work or school, to resume normal activities, and for skin to completely heal. The median time to return to work or school was 2 days (IQR, 1-7 days), to resuming normal activities was 10 days (IQR, 3-14 days), and for skin to completely heal was 30 days (IQR, 14-45 days). Table 3 presents 3 reduced mixed-effects, multivariate-adjusted coefficients of patient and surgical characteristics associated with postsurgical recovery metrics. A greater number of estimated days of missed school or work was significantly associated with surgical procedures in the inframammary region (1.2 days; 95% CI, −0.9 to 3.3 days; β, 0.26; 95% CI, 0.11-0.40; P = .002) compared with procedures in the axilla. Tumor necrosis factor use, a likely marker of disease severity, was significantly associated with a shorter recovery time (−0.9 days; 95% CI, −3.0 to 1.2 days; β, −0.21; 95% CI, −0.33 to −0.08; P < .001).
Table 3. Mixed-Effects Multiple Linear Regression of Patient-Reported Recovery Associated With Patient Demographics, Clinical Risk Factors, and Surgery Characteristicsa.
| Characteristic | Days of missed work or school during recoveryb | Days to return to normal activityc | Days for skin to completely heal overd | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Sample median (IQR) | Estimated change (95% CI), d | Adjusted coefficient (95% CI)e | Sample median (IQR) | Estimated change (95% CI), d | Adjusted coefficient (95% CI)f | Sample median (IQR) | Estimated change (95% CI), d | Adjusted coefficient (95% CI)g | |
| Race and ethnicity | |||||||||
| Non-White | 3 (1 to 10) | NA | NA | 13 (7 to 15) | NA | NA | 30 (21 to 60) | 11.4 (11.4 to 11.4)h | 1.14 (0.20 to 2.08)h |
| White | 2 (0 to 4.5) | NA | NA | 5 (0 to 14) | NA | NA | 15 (10 to 30) | 1 [Reference] | 1 [Reference] |
| TNF inhibitor | 3 (1 to 12) | −0.9 (−3.0 to 1.2) | −0.21 (−0.33 to −0.08)h | 12 (3 to 18) | NA | NA | 25 (14 to 45) | NA | NA |
| Surgery type | |||||||||
| Deroofing | 5 (0 to 14) | NA | NA | 7 (3 to 14) | 1 [Reference] | 1 [Reference] | 30 (14 to 56) | 1 [Reference] | 1 [Reference] |
| Local excision with closure | 2 (1 to 5) | NA | NA | 10 (7 to 14) | 0.5 (−1.6 to 2.6)h | 0.07 (0.01 to 0.12)h | 25 (14 to 30) | 0.9 (0.9 to 0.9)h | 0.08 (0.03 to 0.14)h |
| Local excision without closure | 1 (1 to 7) | NA | NA | 7 (2 to 14) | 0.4 (−1.7 to 2.6) | 0.05 (−0.07 to 0.18) | 42 (28 to 90) | 0.8 (0.8 to 0.8) | 0.08 (−0.04 to 0.20) |
| Surgical location | |||||||||
| Axilla | 4 (0 to 14) | 1 [Reference] | 1 [Reference] | 7 (5 to 14) | 1 [Reference] | 1 [Reference] | 30 (16 to 42) | NA | NA |
| Inframammary | 3 (1 to 10) | 1.2 (−0.9 to 3.3) | 0.26 (0.11 to 0.40)h | 7 (3 to 11) | 0.7 (−1.4 to 2.7)h | 0.09 (0.02 to 0.15)h | 17.5 (14 to 30) | NA | NA |
| Inguinal or pubic region | 2 (1 to 5) | 0.1 (−2.0 to 2.1) | 0.01 (−0.13 to 0.16) | 14 (4 to 14) | 1.0 (−2.0 to 2.1) | 0.01 (−0.07 to 0.08) | 25 (15 to 40) | NA | NA |
| Labia, scrotum, or perianal region or perineum | 2 (1 to 14) | 0.1 (−2.0 to 2.2) | 0.02 (−0.15 to 0.19) | 7 (2 to 18) | 0.0 (−2.0 to 2.1) | 0.00 (−0.09 to 0.09) | 44 (10 to 90) | NA | NA |
| Buttock, thigh, neck, or other | 2 (1 to 7) | 0.2 (−1.9 to 2.2) | 0.05 (−0.10 to 0.19) | 10 (3 to 16) | 0.1 (−1.9 to 2.2) | 0.02 (−0.05 to 0.08) | 30 (20 to 56) | NA | NA |
Abbreviations: NA, not applicable; TNF, tumor necrosis factor.
Only significant variables in the final reduced model are shown. The initial fully adjusted model included age group, sex, race and ethnicity, body mass index group, current smoking status, diabetes, TNF inhibitor use, surgery type, and surgical location as covariates.
The median in the sample was 2 days (IQR, 1-7 days).
The median in the sample was 10 days (IQR, 3-14 days).
The median in the sample was 30 days (IQR, 14-45 days).
The reduced mixed-effects (random effects for patient) model adjusted for TNF inhibitor use and surgical location. There was no difference in estimated days missed from work or school (natural log transformed) compared with the fully adjusted model (likelihood ratio test, P = .82; n = 102).
The reduced mixed-effects (random effects for patient) model adjusted for surgery type and surgery location. There was no difference in estimated days to return to normal activity (natural log transformed) compared with the fully adjusted model (likelihood ratio test, P = .51; n = 125).
The reduced mixed-effects (random effects for patient) model adjusted for race and ethnicity and surgery type. There was no difference in estimated days for skin to completely heal (square root transformed) compared with the fully adjusted model (likelihood ratio test, P = .21; n = 142).
Significant at P < .05.
A greater number of estimated days to return to normal activity was significantly associated with local excision with closure (0.5 days; 95% CI, −1.6 to 2.6 days; β, 0.07; 95% CI, 0.01-0.12; P = .02) compared with deroofing and with surgical procedures in the inframammary region (0.7 days; 95% CI, −1.4 to 2.7 days; β, 0.09; 95% CI, 0.02-0.15; P = .01) compared with the axilla. Longer healing time was significantly associated with non-White race and ethnicity (11.4 days; 95% CI, 11.4-11.4 days; β, 1.14; 95% CI, 0.20-2.08; P = .02) compared with White race and ethnicity and with local excision with closure (0.9 days; 95% CI, 0.9-0.9 days; β, 0.08; 95% CI, 0.03-0.14; P = .003) compared with deroofing.
Although diabetes, smoking, and morbid obesity are typically considered risk factors for poor healing,35 these variables were not significantly associated with recovery metrics in the present sample. In addition, age group and sex were not significantly associated with recovery.
Patient Satisfaction With Surgical Experience and Outcomes
Most patients reported that they would recommend the surgery to a friend (163 of 194 procedures [84%]), have the surgery again (162 procedures [84%]), were glad they underwent the surgery (166 procedures [86%]), and believed the surgery met their expectations (158 procedures [81%]) (Figure, A). Because of the limited number of patients reporting dissatisfaction, a multivariate logistic regression model could not be used to help identify adjusted variables associated with survey responses. Instead, a series of bivariate logistic regression analyses was performed (Table 4). Age greater than 35 years (OR, 4.93; 95% CI, 1.45-16.79; P = .01) and obesity class 3 (body mass index [calculated as weight in kilograms divided by height in meters squared], ≥40) (OR, 4.48; 95% CI, 1.21-16.52; P = .02) were associated with higher odds of satisfaction with the surgical procedure outcomes. Current smoking status (OR, 0.14; 95% CI, 0.05-0.34; P < .001), local excision without closure (OR, 0.21; 95% CI, 0.06-0.75; P = .02), inguinal or pubic surgical location (OR, 0.31; 95% CI, 0.10-0.99; P = .048), and recurrence (OR, 0.33; 95% CI, 0.14-0.80; P = .01) were associated with lower odds of satisfaction. Given possible recall bias related to time since last surgery or bias related to multiple procedures, time between the most recent surgery and questionnaire completion and history of procedures were tested against satisfaction with the surgical procedure and were found to have no association.
Figure. Patient-Reported Satisfaction With the Surgical Procedure.
HS indicates hidradenitis suppurativa.
Table 4. Bivariate Logistic Regression of Characteristics Associated With Patient-Reported Satisfaction.
| Characteristic | Bivariate OR (95% CI) | |
|---|---|---|
| Satisfaction with procedure | Satisfaction with outcome | |
| Age at time of surgery, y | ||
| <25 | 1 [Reference] | 1 [Reference] |
| 25-35 | 1.45 (0.56-3.75) | 1.85 (0.76-4.51) |
| >35 | 4.93 (1.45-16.79)a | 9.43 (2.52-35.29)a |
| Sex | ||
| Female | NAb | 1 [Reference] |
| Male | NAb | 4.69 (1.07-20.57)a |
| Race and ethnicityc | ||
| Non-Whited | 0.90 (0.37-2.18) | 0.41 (0.16-1.08) |
| White | 1 [Reference] | 1 [Reference] |
| BMIe | ||
| Normal | 1 [Reference] | 1 [Reference] |
| Overweight | 0.69 (0.18-2.68) | 0.96 (0.20-4.56) |
| Obesity class 1 | NAf | 5.56 (1.29-24.10)a |
| Obesity class 2 | 0.71 (0.22-2.33) | 0.94 (0.25-3.58) |
| Obesity class 3 | 4.48 (1.21-16.52)a | 0.95 (0.31-2.90) |
| Current smoker | 0.14 (0.05-0.34)a | 0.65 (0.24-1.76) |
| Diabetes | 0.90 (0.38-2.17) | 1.65 (0.66-4.12) |
| TNF inhibitor | 1.21 (0.52-2.79) | 0.46 (0.20-1.08) |
| Surgery type | ||
| Deroofing | 1 [Reference] | 1 [Reference] |
| Local excision with closure | 2.53 (0.71-9.02) | 0.91 (0.37-2.25) |
| Local excision without closure | 0.21 (0.06-0.75)a | 0.81 (0.16-4.01) |
| Surgical location | ||
| Axilla | 1 [Reference] | 1 [Reference] |
| Inframammary | 0.59 (0.13-2.70) | 0.69 (0.23-2.11) |
| Inguinal or pubic region | 0.31 (0.10-0.99)a | 1.93 (0.23-5.94) |
| Labia, scrotum, or perianal region or perineum | 1.53 (0.17-13.92) | NAg |
| Buttock, thigh, neck, or other | 0.48 (0.14-1.69) | 1.62 (0.52-4.99) |
| Recurrence | 0.33 (0.14-0.80)a | 0.39 (0.17-0.88)a |
| Time between surgery and questionnaire | 0.99 (0.96-1.02) | 0.98 (0.95-1.01) |
| Prior surgeries | 2.13 (0.91-4.97) | 0.82 (0.34-1.91) |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); NA, not applicable; OR, odds ratio; TNF, tumor necrosis factor.
Significant at P < .05.
Unable to calculate because all male patients reported satisfaction with surgery.
Unable to calculate values for unknown race because all patients of unknown race reported satisfaction with surgery. This information was retrieved from electronic medical records. The information was self-reported, and patients had the option to not report certain items.
The non-White category included the following racial and ethnic groups: Black, American Indian or Alaska Native, Asian (including Indian), and Native Hawaiian or Other Pacific Islander.
Normal BMI was considered 20.0 to 24.9; overweight, 25.0 to 29.9; obesity class 1, 30.0 to 34.9; obesity class 2, 35.0 to 39.9; and obesity class 3, 40 or greater.
Unable to calculate because all patients with a BMI obesity class 1 reported satisfaction with surgery.
Unable to calculate because all observations related to this location included reports of satisfaction with outcome.
After surgery, most patients were either extremely or very satisfied with procedure results (148 of 194 procedures [76%]), appearance of the healed wound (112 procedures [58%]), and recovery time (143 procedures [74%]) (Figure, B). In a series of bivariate logistic regressions, age greater than 35 years (OR, 9.43; 95% CI, 2.52-35.29; P = .001), male sex (OR, 4.69; 95% CI, 1.07-20.57; P = .04), and obesity class 1 (OR, 5.56; 95% CI, 1.29-24.10; P = .02) were associated with a greater satisfaction. Patients with recurrence were less likely to be satisfied with procedure outcomes (OR, 0.39; 95% CI, 0.17-0.88; P = .02).
When asked to rate intraoperative pain on a scale of 0 to 10, where 10 represented the worst pain, for 61 of 194 procedures (31%), patients reported a score of 0 (no pain at all), and for 87 procedures (45%), patients reported a score between 1 and 4 (Figure, C). During the first week of recovery after surgery, for 53 procedures (27%), patients reported a pain level of 5 (moderate pain) or lower. Although for 63 procedures (32%), patients reported a pain level of 8 or greater during their first week of recovery, for 126 procedures (65%), patients reported that they believed the pain experienced during an HS flare was worse than their surgical recovery pain.
Discussion
In this cohort study, most of the reported patient characteristics and outcomes were consistent with and expanded on previous reports.16,17,21,22,23,24 Most patients in the present study were female and in the non-White race and ethnicity category, which is more reflective of the population with HS than most prior reports of HS surgical outcomes.16,17,21,22,23,24,25,27,28 Never smoking and Hurley stage 3 disease were also more common in the present study than in most previous studies.16,22,26,28
The recurrence rate (41%) was higher than those reported in some other studies (<5%-25%).16,21,36 Response rates for patients seen in the clinic for follow-up were higher than for those contacted by telephone, and it is likely that patients returning to the clinic over time were more likely to have had recurrences than were those who did not return for follow-up and could not be reached by telephone. Recall bias also likely influenced results because some subjects may not have accurately remembered the timing or location of transient flares relative to the timing of procedures. During survey administration, many patients who had undergone multiple procedures also commented that they could not recall which procedures were specifically followed by recurrences; thus, they reported recurrence for all procedures. It is possible that this type of recall bias led to overestimation of recurrences. It is also possible that interpretation of questions about pain varied among patients because pain was assessed by asking “how much pain was experienced” during and after the procedure, but maximal and average pain levels were not specified. As in previous studies, recurrence risk decreased in association with increasing age.21 This finding was likely attributable to disease tending to be most progressive in the third and fourth decades of life and often stabilizing in later years. Despite recurrences, patient satisfaction with procedures and outcomes remained high. In our experience, recurrences often result in more temporary or limited disease activity than that encountered before surgery, leading to overall benefit. In the present study, recurrence alone appeared to not be a marker of treatment failure given that satisfaction remained high.
The median number of days to return to work or school was 2 (IQR, 1-7 days), and the median number of days to return to normal activity was 10 (IQR, 3-14 days), which was shorter than in another study reporting a mean of 1 to 4 weeks.20 Diabetes, smoking, and obesity were not associated with more days missed from work or school, days to return to normal activity, or days until complete healing. This finding suggests that diabetes, smoking, and obesity have, at most, a modest correlation with wound healing after HS surgery and are not contraindications if patients are aware of the potential risks. Based on results in the overall cohort, it may be reasonable to counsel patients that the typical time away from work or school after clinic-based excision or deroofing is less than 1 week, return to normal activity is less than 2 weeks, and typical complete healing occurs in 4 to 5 weeks.
Most patients were satisfied or very satisfied with surgical outcomes. Current smoking was associated with an increased risk of dissatisfaction, although most patients were still satisfied. Compared with deroofing, local excision without closure was associated with higher odds of recurrence and lower satisfaction with the procedure. Bias likely existed in selecting deroofing or excision based on the size and location of affected areas or demographic factors that could have influenced recurrence. The decision to perform deroofing vs excision was often influenced by patient preference for managing closed vs open wounds; thus, these preferences may have been associated with differences in patient expectations and satisfaction. Excision may also be avoided in certain areas, such as large wounds in the crura, where suturing may result in more restrictive scarring. Some of the largest wounds may have been more challenging to close primarily because of their size and shape; thus, there may have been a tendency to leave more complicated wounds to heal by secondary intention after deroofing rather than by complete excision with more challenging reconstructions. A randomized clinical trial comparing local excision with deroofing is essential to understanding an optimal surgical approach in the clinical setting.
Although positive outcomes are crucial to judging surgical success, the morbidity associated with surgery and recovery is also important to understand in balancing risks and benefits. Most patients reported mild pain during the surgical procedure and mild to moderate pain during the first week of recovery. More than two-thirds (77%) found recovery pain to be the same as or less severe than typical HS-related pain during disease flares. This finding is important to consider because many patients and clinicians may hesitate to pursue surgery because of fear of discomfort, but most patients reported that the benefits outweighed the relatively tolerable procedure and recovery.
Although we analyzed several variables in a single analysis for some outcomes, we did not apply the Bonferroni correction. This is an area of controversy because the Bonferroni correction is often believed to be overly aggressive and to decrease the sensitivity for identifying relevant associations.31,32,33,34 Given the exploratory nature of our analysis, we ultimately believed that the benefit of maintaining appropriate sensitivity outweighed the drawbacks of potential overcorrection.
Limitations
This study was limited to a single institution with procedures performed by a single surgeon in a subspecialty HS clinic. Responses were only recorded from 78 of 201 eligible patients; thus, the potential for reporting bias was high. However, a relatively large number of patients and procedural experiences were included in the analysis and provided data that enhanced the current understanding of patient outcomes and experiences after clinic-based HS surgery. Given the retrospective nature of data collection and potential confusion related to remembering multiple procedures, there was risk of recall bias. An additional limitation is that all patients were recruited from a single subspecialty clinic, which may not reflect the general population of individuals with HS, and the multiple analyses performed may have contributed to some borderline significant P values being attributable to chance alone. Future studies using prospective data collection and randomized trial designs comparing interventions are critical to understanding how we can optimize surgical interventions for HS in the future.
Conclusions
In this cohort study, patients reported high rates of satisfaction with clinic-based HS surgery. Recovery was typically rapid, with most patients rating postsurgical pain as less severe than their HS pain. Smoking status, obesity, and diabetes were not significantly associated with speed of recovery and wound healing.
References
- 1.Ather S, Chan DSY, Leaper DJ, Harding KG. Surgical treatment of hidradenitis suppurativa: case series and review of the literature. Int Wound J. 2006;3(3):159-169. doi: 10.1111/j.1742-481X.2006.00235.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kurzen H, Kurokawa I, Jemec GBE, et al. What causes hidradenitis suppurativa? Exp Dermatol. 2008;17(5):455-456. doi: 10.1111/j.1600-0625.2008.00712.x [DOI] [PubMed] [Google Scholar]
- 3.Revuz JE, Canoui-Poitrine F, Wolkenstein P, et al. Prevalence and factors associated with hidradenitis suppurativa: results from two case-control studies. J Am Acad Dermatol. 2008;59(4):596-601. doi: 10.1016/j.jaad.2008.06.020 [DOI] [PubMed] [Google Scholar]
- 4.Jemec GBE, Kimball AB. Hidradenitis suppurativa: epidemiology and scope of the problem. J Am Acad Dermatol. 2015;73(5)(suppl 1):S4-S7. doi: 10.1016/j.jaad.2015.07.052 [DOI] [PubMed] [Google Scholar]
- 5.Miller IM, McAndrew RJ, Hamzavi I. Prevalence, risk factors, and comorbidities of hidradenitis suppurativa. Dermatol Clin. 2016;34(1):7-16. doi: 10.1016/j.det.2015.08.002 [DOI] [PubMed] [Google Scholar]
- 6.Reeder VJ, Mahan MG, Hamzavi IH. Ethnicity and hidradenitis suppurativa. J Invest Dermatol. 2014;134(11):2842-2843. doi: 10.1038/jid.2014.220 [DOI] [PubMed] [Google Scholar]
- 7.Wertenteil S, Strunk A, Garg A. Association of low socioeconomic status with hidradenitis suppurativa in the United States. JAMA Dermatol. 2018;154(9):1086-1088. doi: 10.1001/jamadermatol.2018.2117 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Clemmensen OJ. Topical treatment of hidradenitis suppurativa with clindamycin. Int J Dermatol. 1983;22(5):325-328. doi: 10.1111/j.1365-4362.1983.tb02150.x [DOI] [PubMed] [Google Scholar]
- 9.Gener G, Canoui-Poitrine F, Revuz JE, et al. Combination therapy with clindamycin and rifampicin for hidradenitis suppurativa: a series of 116 consecutive patients. Dermatology. 2009;219(2):148-154. doi: 10.1159/000228334 [DOI] [PubMed] [Google Scholar]
- 10.Mendonça CO, Griffiths CEM. Clindamycin and rifampicin combination therapy for hidradenitis suppurativa. Br J Dermatol. 2006;154(5):977-978. doi: 10.1111/j.1365-2133.2006.07155.x [DOI] [PubMed] [Google Scholar]
- 11.Jemec GB, Wendelboe P. Topical clindamycin versus systemic tetracycline in the treatment of hidradenitis suppurativa. J Am Acad Dermatol. 1998;39(6):971-974. doi: 10.1016/S0190-9622(98)70272-5 [DOI] [PubMed] [Google Scholar]
- 12.Haslund P, Lee RA, Jemec GBE. Treatment of hidradenitis suppurativa with tumour necrosis factor-alpha inhibitors. Acta Derm Venereol. 2009;89(6):595-600. doi: 10.2340/00015555-0747 [DOI] [PubMed] [Google Scholar]
- 13.Yamauchi PS, Mau N. Hidradenitis suppurativa managed with adalimumab. J Drugs Dermatol. 2009;8(2):181-183. [PubMed] [Google Scholar]
- 14.Delage M, Samimi M, Atlan M, Machet L, Lorette G, Maruani A. Efficacy of infliximab for hidradenitis suppurativa: assessment of clinical and biological inflammatory markers. Acta Derm Venereol. 2011;91(2):169-171. doi: 10.2340/00015555-1025 [DOI] [PubMed] [Google Scholar]
- 15.Tursi A. Concomitant hidradenitis suppurativa and pyostomatitis vegetans in silent ulcerative colitis successfully treated with golimumab. Dig Liver Dis. 2016;48(12):1511-1512. doi: 10.1016/j.dld.2016.09.010 [DOI] [PubMed] [Google Scholar]
- 16.van der Zee HH, Prens EP, Boer J. Deroofing: a tissue-saving surgical technique for the treatment of mild to moderate hidradenitis suppurativa lesions. J Am Acad Dermatol. 2010;63(3):475-480. doi: 10.1016/j.jaad.2009.12.018 [DOI] [PubMed] [Google Scholar]
- 17.Büyükaşik O, Hasdemir AO, Kahramansoy N, Çöl C, Erkol H. Surgical approach to extensive hidradenitis suppurativa. Dermatol Surg. 2011;37(6):835-842. [DOI] [PubMed] [Google Scholar]
- 18.Scuderi N, Monfrecola A, Dessy LA, Fabbrocini G, Megna M, Monfrecola G. Medical and surgical treatment of hidradenitis suppurativa: a review. Skin Appendage Disord. 2017;3(2):95-110. doi: 10.1159/000462979 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Janse I, Bieniek A, Horváth B, Matusiak Ł. Surgical procedures in hidradenitis suppurativa. Dermatol Clin. 2016;34(1):97-109. doi: 10.1016/j.det.2015.08.007 [DOI] [PubMed] [Google Scholar]
- 20.Kohorst JJ, Baum CL, Otley CC, et al. Surgical management of hidradenitis suppurativa: outcomes of 590 consecutive patients. Dermatol Surg. 2016;42(9):1030-1040. doi: 10.1097/DSS.0000000000000806 [DOI] [PubMed] [Google Scholar]
- 21.Kohorst JJ, Baum CL, Otley CC, et al. Patient satisfaction and quality of life following surgery for hidradenitis suppurativa. Dermatol Surg. 2017;43(1):125-133. doi: 10.1097/DSS.0000000000000942 [DOI] [PubMed] [Google Scholar]
- 22.Kofler L, Schweinzer K, Heister M, Kohler M, Breuninger H, Häfner HM. Surgical treatment of hidradenitis suppurativa: an analysis of postoperative outcome, cosmetic results and quality of life in 255 patients. J Eur Acad Dermatol Venereol. 2018;32(9):1570-1574. doi: 10.1111/jdv.14892 [DOI] [PubMed] [Google Scholar]
- 23.Fertitta L, Hotz C, Wolkenstein P, et al. Efficacy and satisfaction of surgical treatment for hidradenitis suppurativa. J Eur Acad Dermatol Venereol. 2020;34(4):839-845. doi: 10.1111/jdv.16135 [DOI] [PubMed] [Google Scholar]
- 24.Rompel R, Petres J. Long-term results of wide surgical excision in 106 patients with hidradenitis suppurativa. Dermatol Surg. 2000;26(7):638-643. doi: 10.1046/j.1524-4725.2000.00043.x [DOI] [PubMed] [Google Scholar]
- 25.Ritz JP, Runkel N, Haier J, Buhr HJ. Extent of surgery and recurrence rate of hidradenitis suppurativa. Int J Colorectal Dis. 1998;13(4):164-168. doi: 10.1007/s003840050159 [DOI] [PubMed] [Google Scholar]
- 26.Blok JL, Boersma M, Terra JB, et al. Surgery under general anaesthesia in severe hidradenitis suppurativa: a study of 363 primary operations in 113 patients. J Eur Acad Dermatol Venereol. 2015;29(8):1590-1597. doi: 10.1111/jdv.12952 [DOI] [PubMed] [Google Scholar]
- 27.Jemec GB. Effect of localized surgical excisions in hidradenitis suppurativa. J Am Acad Dermatol. 1988;18(5 Pt 1):1103-1107. doi: 10.1016/S0190-9622(88)70113-9 [DOI] [PubMed] [Google Scholar]
- 28.Deckers IE, Dahi Y, van der Zee HH, Prens EP. Hidradenitis suppurativa treated with wide excision and second intention healing: a meaningful local cure rate after 253 procedures. J Eur Acad Dermatol Venereol. 2018;32(3):459-462. doi: 10.1111/jdv.14770 [DOI] [PubMed] [Google Scholar]
- 29.Vlassova N, Kuhn D, Okoye GA. Hidradenitis suppurativa disproportionately affects African Americans: a single-center retrospective analysis. Acta Derm Venereol. 2015;95(8):990-991. doi: 10.2340/00015555-2176 [DOI] [PubMed] [Google Scholar]
- 30.Vaidya T, Vangipuram R, Alikhan A. Examining the race-specific prevalence of hidradenitis suppurativa at a large academic center: results from a retrospective chart review. Dermatol Online J. 2017;23(6):13030/qt9xc0n0z1. doi: 10.5070/D3236035391 [DOI] [PubMed] [Google Scholar]
- 31.Perneger TV. What’s wrong with Bonferroni adjustments. BMJ. 1998;316(7139):1236-1238. doi: 10.1136/bmj.316.7139.1236 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Perneger TV. Adjusting for multiple testing in studies is less important than other concerns. BMJ. 1999;318(7193):1288. doi: 10.1136/bmj.318.7193.1288a [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Aickin M. Other method for adjustment of multiple testing exists. BMJ. 1999;318(7176):127-128. doi: 10.1136/bmj.318.7176.127a [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Bender R, Lange S. Multiple test procedures other than Bonferroni’s deserve wider use. BMJ. 1999;318(7183):600-601. doi: 10.1136/bmj.318.7183.600a [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Sørensen LT. Wound healing and infection in surgery: the pathophysiological impact of smoking, smoking cessation, and nicotine replacement therapy: a systematic review. Ann Surg. 2012;255(6):1069-1079. doi: 10.1097/SLA.0b013e31824f632d [DOI] [PubMed] [Google Scholar]
- 36.Bohn J, Svensson H. Surgical treatment of hidradenitis suppurativa. Scand J Plast Reconstr Surg Hand Surg. 2001;35(3):305-309. doi: 10.1080/028443101750523230 [DOI] [PubMed] [Google Scholar]