Introduction

Different management strategies exist for patients with ventral hernias including operative repair and non-operative management. There are a variety of reasons patients or clinicians choose non-operative management including prohibitive medical risks, high rate of predicted surgical complications, patient preference, or asymptomatic hernias [1–14].

The natural history of ventral hernias managed non-operatively is understudied and largely unknown. In a related but different disease process, inguinal hernias, the risk of acute presentation is low (<1%/patient year). However, based on a randomized trial with at least 7 years of follow-up, the risk of crossover to elective surgery is high (>68% of patients at 10 years) due to progression of size and symptoms [15, 16]. Only small studies or those with limited methodology (e.g., retrospective or substantial loss to follow-up) have been reported on patients with ventral hernias [17–20]. The rate of acute presentation, emergency room visits, and progression of symptoms for ventral hernias treated non-operatively is unknown.

The aim of this study is to determine clinical and patient-reported outcomes of patients who undergo non-operative treatment of a ventral hernia. We hypothesize that (1) the annual risk of emergency ventral hernia repair is low (<1%/year), (2) the rate of conversion to elective ventral hernia repair is high (>20%/year), and (3) patients managed non-operatively will experience a decline in quality of life and functional status.

Methods

Following Institutional Review Board (IRB) approval and registration on Clinicaltrials.gov (NCT02457364), patients with a ventral hernia to be treated with non-operative management were enrolled in a prospective study from 2014 to 2015. Patients were recruited through a hernia clinic at a safety-net hospital, which provides care for low-income, underserved, and uninsured populations. In a recent study performed at the same hernia clinic, the population was comprised largely of unemployed (77%), uninsured (81%), and minority (88%) patients [21]. As described in previous studies, the general patient population seen in this clinic were 43% male, 19% smokers, with an average BMI of 34.1, [22] Hernias were 30% primary and 70% incisional, while 17% were recurrent [22]. Average hernia width was 4.5 cm measured by computed tomography [22]. The vast majority of these clinic patients are referred either from primary care clinics or the emergency rooms that are both within the same safety-net healthcare system.

Inclusion criteria were any patient undergoing initial non-operative management determined by surgeon judgment or patient choice. In this clinic, non-operative management is generally chosen by the surgeon for BMI >33 kg/m2, hemoglobin A1c >8%, current smokers, and end-stage high-risk medical comorbidities (e.g., advanced cardiopulmonary disease) [23, 24]. Exclusion criteria were any patient unlikely to be able to be able to follow-up (e.g., patients with plans on moving out of the region) and patients excluded by the IRB as “at-risk” (e.g., prisoners).

Baseline data recorded included patient demographic data, comorbidities, hernia data, and reason(s) for non-operative management. Patient demographic data recorded included age, gender, and ethnicity. Patient comorbidities recorded were American Society of Anesthesiologists (ASA) score, BMI, smoking, chronic obstructive pulmonary disease (COPD), diabetes mellitus (DM), immunosuppression, prostate disease, aneurysmal disease, prior surgical site infection (SSI), ascites, and number of prior abdominal surgeries. Trained abstractors determined ASA based on comorbidities at the time of initial consultation. BMI was measured at the time of initial clinic visit. Smoking was defined as use of any cigarettes on a daily basis. COPD was determined by any documentation of COPD in the medical record. DM was defined as hemoglobin A1c >6.5% or use of any diabetic medications. Immunosuppression was defined as any immunosuppressive medication use, chemotherapy within 1 month, or immunosuppressive disease. Prostate disease was defined as any diagnosis of benign prostatic hypertrophy, prostate cancer, or prostatitis. Prior SSI was determined by documentation of SSI in the EMR or by patient report. Ascites was determined by physical exam or findings of ascites on imaging. The number of prior abdominal surgeries was determined by patient report. Hernia data obtained included hernia type (primary versus incisional), recurrent, history of prior ventral hernia repair, hernia size (area), and hernia location (medial or lateral to semilunar line [25]). Hernia size was determined by CT measurements and supplemented with physical exam measurements if CT was not available.

Patient-reported outcomes were obtained through patient surveys given prior to surgeon consultation and at each follow-up (Table 1). Patients were asked if they wanted their hernia repaired and why. Patients were allowed to list one or more reasons that they desired hernia repair. They were asked to rate satisfaction and cosmetic satisfaction with their abdomen (1–10 Likert scale, 1 = dissatisfied, 10 = satisfied), abdominal pain (Visual Analog Scale, 1 = no pain 10 = severe pain), and patient function (using a modified Activities Assessment Scale, AAS, 1 = poor function and 10 = good function). An overall function score was calculated as the sum of all of the function questions normalized to 100 (1 = poor function and 100 = good function) [26, 27].

Table 1 Patient survey
Full size table

Patients were contacted one year following initial enrollment. On follow-up, patients were asked about hernia-related emergency room visits since enrollment, abdominal surgery, hernia repair (elective or emergent), current ventral hernia status, and the previously described patient-reported outcomes.

The primary outcome was rate of surgical repair. Secondary outcomes included rate of emergency repair, rate of elective repair, number of emergency room visits, and change in patient-reported outcomes. The risk of emergency repair was defined as low if <1%/year, moderate if 1–5%/year, and high if >5%/year. The risk of crossover to surgical repair was defined as low if <5%/year, moderate if 5–10%/year, and high if >10%/year. These were derived from prior non-operative studies for inguinal hernia and ventral hernia [15, 23, 28, 29]. Outcomes of patients who continued non-operative management were analyzed separately from patients who crossed-over to operative management. Initial and follow-up scores were reported as median (interquartile range), and differences were compared using a Wilcoxon matched-pairs signed rank test.

Hypothesis-generating post hoc subgroup analysis by hernia type and by hernia size was performed. Hernia type was defined as primary versus incisional [25]. Hernia size was broken down into small (<4 cm), medium (≥4–10 cm) and large (≥10 cm) based on hernia width [25].

Results

A total of 128 patients who underwent non-operative management of a ventral hernia were enrolled, and 99 were available for follow-up. Patient demographics, comorbidities, and hernia data are shown in Table 2. Patients with no follow-up were more likely to be male, smokers, had fewer prior ventral hernia repairs and fewer prior abdominal surgeries, and were more likely to have had a primary ventral hernia. Though it was not statistically significant, there was a trend toward higher ASA score in patients lost to follow-up compared to those who were followed. Of the 128 patients, reasons for non-operative management were obesity (n = 87, 68.0%), smoking (n = 23, 18.0%), other comorbidities (n = 20, 15.6%), patient choice (n = 12, 9.4%), and surgical complexity (n = 2, 1.6%). Most patients (n = 100, 78.1%) desired surgical repair of their hernia. Reasons for wanting surgery were pain (n = 89, 69.5%), cosmesis (n = 23, 18.0%), functional limitations (n = 18, 14.1%), desire for repair of enterocutaneous fistulas (2, 1.6%), desire for concomitant ostomy reversal (5, 3.9%), itching (1, 0.8%), nausea (2, 1.6%), fear of incarceration/complications (4, 3.1%), and chronic infected mesh (1, 0.8%).

Table 2 Patient demographics, comorbidities, and hernia data (N = 128)
Full size table

Follow-up was obtained for 99 (77.3%) patients with a median (interquartile range) of 12.2 (10.4–13.5) months. Twenty (20.2%) patients had at least 1 emergency room visit related to their hernia. Of the 99 patients who were followed, 23 (23.2%) underwent operative management, 3 (3.0%) emergently and 20 (20.2%) electively. The 3 emergent ventral hernia repairs were for obstruction (n = 2) and for an unknown reason at an outside institution (n = 1). None of these were found to have strangulation of bowel. Of the 20 who underwent elective repair, 13 of the patients met goals required for elective surgery (e.g., weight loss, quitting smoking), 4 had surgery at outside institutions, 2 had elective repair due to chronic incarceration and pain, and 1 had elective repair after coming the emergency room multiple times for the hernia. Of all 23 patients who ultimately underwent surgery, there were 3 surgical site infections, 2 seromas, 2 wound dehiscences, and 1 hernia recurrence.

Based on the patient-reported outcome surveys, patients who were managed non-operatively had no change in their scores on any of the survey questions; however, patients managed operatively had substantial improvements in their scores (Fig. 1; Table 3).

Fig. 1
figure 1

Initial and follow-up survey scores for non-operative patients and patients converted to operative management (1 = poor satisfaction 10 = good satisfaction, 1 = no pain 1 = severe pain, 1 = poor function 100 = good function)

Full size image
Table 3 Number of patients with improved, no change, or worsened patient-reported outcomes
Full size table

On subgroup analysis, patients with primary and incisional hernias had similar rates of ER visits, surgical repair, and emergent surgical repair (Table 4). There was a trend toward more ER visits, surgical repairs, and emergent repairs as hernia size increased; however, this was not significant due to insufficient power (Table 4).

Table 4 Subgroup analysis by hernia type and hernia size
Full size table

Discussion

Among patients with ventral hernias initially managed non-operatively, the short-term risk of emergency surgery is moderate (3% at 1 year); thus, non-operative management appears to be safe in the short term. However, these patients frequently have other problems associated with their hernia. Specifically, 20.2% of patients with at least 1 emergency room visit related to their hernia, and 4% sought surgery elsewhere. Furthermore, most ventral hernia patients experience poor satisfaction with their abdominal wall, with a median satisfaction score of 2 out of 10 (where 10 = high satisfaction). It remains to be seen if these risks are cumulative over time.

The majority of patients (79.8%, 76/99) in this study were managed non-operatively due to potentially modifiable risk factors such as obesity and smoking. However, only 17.1% (13/76) of these patients were able to modify their personal risk factors enough to qualify for elective surgery. Presently, there is no widely accepted management strategy nationally for patients with modifiable risk factors; however, institutionally, there is a policy to avoid elective repair on obese patients and patients who are active smokers. Some patients initially managed non-operatively simply went to another institution to seek elective surgery. Given these challenges, national consensus guidelines on the management of patients with potentially modifiable risk factors and evaluation of preoperative risk-reduction programs in these patients are needed [23].

Patients initially managed non-operatively and converted to surgical repair demonstrated a significant improvement in satisfaction and function. In contrast, patients who remained non-operative had little change in their patient-reported outcomes and functional status. This suggests that operative management of ventral hernias may improve patient satisfaction and function compared to non-operative treatment. However, long-term follow-up is needed to see if the improved patient-reported outcomes are durable, particularly in the face of recurrence or other long-term complications. Furthermore, delaying surgery did not appear to affect outcomes, as there were no cases of bowel strangulation, and few cases of wound complications.

There are several limitations to this study. This study followed patients for a median of 12.2 months, and longer follow-up is needed. This study is scheduled for 5-year follow-up. In addition, this study was only powered to assess the safety and effectiveness of non-operative management and not to compare outcomes of operative versus non-operative management. We chose not to perform a randomized controlled trial because we did not feel that elective surgery on high-risk patients with modifiable comorbidities and substantial predicted complication rates was ethical without an attempt at risk-modification (Clinicaltrials.gov NCT02365194). However, even with this limited power, it appears that operative management may improve patient satisfaction. Furthermore, nearly one-quarter of patients were lost to follow-up, and there were some differences in those followed compared to those lost to follow-up. For example, more male patients were lost than female patients. However, the inclusion of these patients could only have revealed more emergency room visits or elective and emergent hernia repairs (3.0% vs. potential range of 2.3–25% if all patients had followed up), which may have argued against the safety of non-operative management. In addition, phone call follow-up was used for some patients who could not be seen in clinic. This may bias the results through recall and information bias. Next, the results of this study may not be generalizable to other populations. This was performed at a safety-net hospital whose population differs considerably from other hospital populations. This particular population tends to have a high rate of patients who are lost to follow-up, which was problematic in this study. Application of these results to other patient populations should be done cautiously. Finally, this study included all types of hernias, and it is possible that more complex hernias may have different results than simple hernias. We performed a subgroup analysis by hernia type and hernia size to demonstrate possible differences between these; however, these results are hypothesis generating, as this study was not adequately powered or designed to draw definitive conclusions from these subgroups. It is very possible that results for certain subsets of hernias may differ, and further study in this is warranted.

Conclusions

The short-term risks of requiring emergency surgery following non-operative management of a ventral hernia are moderate (3%/1 year), suggesting that there may be a safe window for risk-reduction and preoperative optimization in high-risk patients. Compared to hernia repair that substantially improve patient-reported outcomes, non-operative management was associated with no change in patient satisfaction and function.