All searches were conducted by a health librarian/medical information specialist. Using the Ovid® search interface, the following electronic databases were searched for studies reporting clinical and HRQoL outcomes: MEDLINE® and Epub Ahead of Print, In-Process, In-Data-Review & Other Non-Indexed Citations, Daily and Versions, Embase (1974–2024), EBM Reviews - Cochrane Central Register of Controlled Trials (1991-present), and EBM Reviews - Cochrane Database of Systematic Reviews (2005-present). Searches were performed on 1 August 2024. Search results were limited to records that were published after 2013 to align with the publication of the first Phasix™ Mesh studies [23, 24]. The full search strategy is provided in Supplementary Material 1.

Studies were included if they reported clinical or HRQoL outcomes for patients undergoing open ventral hernia repair using either Phasix™ Mesh or permanent synthetic mesh. The predefined PICOS criteria are outlined in Table 1. If the study methods did not specify whether open, laparoscopic, or robotic-assisted surgical technique was used, it was assumed to be open surgery and included in the SLR. Studies that did not fulfill the inclusion criteria were excluded, with the reasons for exclusion documented. Additional filters were applied (i.e., only full text articles, published 2018 or later, minimum 12-month follow-up, and light/medium weight mesh) post-screening to increase homogeneity across included articles. Articles that passed this stage were further confirmed for inclusion during data extraction, which involved evaluating the study design, baseline characteristics of the study population, and reported efficacy outcomes.

Study selection was conducted using the systematic review platform DistillerSR (DistillerSR Inc., 2021, Ottawa, Canada). Two reviewers independently evaluated the study records, including titles and abstracts, to determine their eligibility. Reviewers recorded the reasons for excluding studies, and any disagreements between the two reviewers were resolved either through consensus or by consulting a third independent reviewer.

Data extraction was conducted by a primary reviewer and independently confirmed by a second reviewer to ensure accuracy. Any disagreements between the two were resolved through discussion, or if necessary, by consulting a third independent reviewer. Data from the selected studies were gathered using standardized extraction templates created in Microsoft® Excel. The extracted study characteristics and outcomes included study information (i.e., author name, publication year, registration number, country of origin, follow-up duration, and study design), baseline population characteristics, treatment and disease details, and efficacy.

Risk of bias for the included studies was assessed using the Newcastle-Ottawa Scale (NOS) as it is specifically designed for observational studies (which comprised the majority of included studies) [35]. The assessment was completed in duplicate by two independent reviewers and any differences in judgments were resolved through discussion, or, if needed, by involving a third reviewer. With the NOS, the total score ranges from 0 to 9 stars, with studies scoring 7 or more considered high quality [35].

Although the SLR included studies reporting both clinical and HRQoL outcomes, HRQoL outcomes were the focus of the study described herein. A narrative synthesis was conducted which included detailed descriptions of study characteristics, methodologies, types of mesh used, and key findings. The interpretation of results considered the study populations, treatment factors, and methodological rigor. Additionally, a narrative overview of each study’s strengths and limitations was provided, along with responses to individual items from the risk of bias assessment to offer insight into the overall quality of the evidence. Quality of life results generated using the Hernia-Related Quality of Life Survey (HerQLes) were prioritized since this tool is validated for use in patients with hernia and results were consistently reported with sufficient detail across studies, thereby minimizing heterogeneity and facilitating robust comparisons.

The database searches yielded 2,944 records; after removing duplicates, 1,770 records were screened at the title and abstract phase, of which 1,306 were excluded. Of the remaining 464 records considered at the full-text screening phase, 297 were excluded for reasons as detailed in the PRISMA flow diagram (Figure 1) [34]. After completion of article screening and applying post-hoc filters to increase homogeneity across the included studies, an additional 98 articles were excluded, ultimately resulting in the inclusion of 69 articles. Of the 69 articles included, 33 reported HRQoL outcomes (with the remainder reporting clinical outcomes).

Of the 33 studies reporting HRQoL outcomes, 28 studies reported such outcomes using 12 different HRQoL scales; namely, abdominal hernia questionnaire (AHQ) [36–41], the Carolinas comfort scale (CCS) [42–44], EuroQol 5-dimension (EQ-5D) [4, 45–47], EuroQol visual analogue scale (EQ VAS) [4], HerQLes [4, 36, 39, 46, 48–50], numeric rating scale (NRS) [42], patient-reported outcomes measurement information system (PROMIS) [50], short-form 12-item survey (SF-12) [43, 51], short-form 36-item survey (SF-36) [52, 53], the visual analogue scale (VAS) [38, 43, 46, 54–60], the ventral hernia screening (VHS) [61], and verbal rating scale (VRS; Table 2) [62]. The remaining five articles reported HRQoL outcomes as the proportion of patients returning to usual activities or with chronic pain and/or swelling [63–67].

Among the studies that reported both pre-operative and post-operative HRQoL scores, all Phasix™ Mesh studies demonstrated improvements in HRQoL, regardless of the assessment instrument used [36, 37, 39–41, 43, 48, 49, 51, 54, 59]. Similarly, studies involving synthetic meshes consistently reported HRQoL improvements, independent of the measurement tool applied [4, 38, 42, 46, 50, 52, 53, 58].

The most commonly used HRQoL instruments across both Phasix™ Mesh and permanent synthetic mesh studies were the AHQ, HerQLes, and VAS (Table 2). Results derived using the AHQ were reported more often in Phasix™ Mesh studies relative to permanent synthetic mesh studies, whereas HerQLes and VAS were reported more evenly across studies of both mesh types (Table 2). The VAS is a validated tool used to measure pain intensity and is not specific to patients with hernia [68], whereas HerQLes is a validated, hernia-specific HRQoL instrument developed specifically for this patient population [69]. Its widespread adoption in hernia research supports its relevance and reliability [30, 70–72]. Therefore, only studies reporting HerQLes results were described further in this narrative review.

In total, seven records reporting HerQLes outcomes were included in this review, comprising six distinct studies and one secondary analysis based on one of the included studies. Four of the included studies evaluated Phasix™ Mesh and three assessed permanent synthetic mesh. The characteristics of the included studies are provided in Table 3. All studies were conducted in the United States and primarily employed retrospective designs, with one randomized controlled trial (RCT) [4] and one secondary analysis of a RCT [46].

Across studies, follow-up durations ranged from approximately 12–66 months, with reported follow-up durations spanning 12.2–66.6 months in Phasix™ Mesh studies and 12–48 months in permanent synthetic mesh studies. This variability in follow-up time points introduced significant heterogeneity and limited the ability to pool outcomes across studies. Sample sizes varied from 51 to 1,203 patients in permanent synthetic mesh studies, and from 51 to 71 patients in Phasix™ Mesh studies.

The mean patient age across studies ranged from the early 50s to mid-60s, and body mass index (BMI) values ranged from approximately 31–33 kg/m². The proportion of male patients ranged from 38% to 57%, and the prevalence of diabetes ranged from 16% to 23.9%.

Hernia disease characteristics of included studies are presented in Table 4. The included studies primarily addressed ventral hernias, with some studies reporting detailed hernia location including umbilical, paramedian, or parastomal hernias. Hernia defect sizes were consistently large, with reported mean or median areas ranging from approximately 273 cm²–345 cm². Varied reporting of hernia dimensions was noted, with lengths typically between 18–23 cm and widths between 13 and 16 cm. The patients included in these studies were complex, as all were diagnosed with large hernias (defined as having a defect of ≥10 cm in width or length, or an area of ≥78.5 cm²); such patients commonly experience high rates of recurrence and complications [73, 74].

HerQLes is a 12-item, 6-point Likert-style survey (with higher scores indicating better results) assessing HRQoL related to abdominal wall function in patients with hernia disease [75]. An increase of at least 15.6 points represents the minimal clinically important difference (MCID) when using the HerQLes [72].

A summary of HerQLes results spanning the included studies is presented in Table 5 and Figure 2. Both Phasix™ Mesh and permanent synthetic mesh studies demonstrated notable improvements in patient-reported HerQLes scores following hernia repair. However, the magnitude and trajectory of improvement varied between mesh types.

In studies evaluating Phasix™ Mesh, improvements in HerQLes scores were observed across multiple follow-up time points relative to baseline. The average increase in HerQLes scores across all time points compared to baseline was approximately 27.5 points. Improvements were evident as early as 0–12 months post-operatively (i.e., mean increase of 25.6 points compared to baseline in Messa 2019 [48] and mean increase of 30.9 points compared to baseline in Christopher, 2021 [49]) and continued to increase through longer-term follow-up, reaching a maximum mean difference of 38.6 points compared to baseline at 36+ months [49]. The longest follow-up timepoint reported was 60 months, whereby a 30.8 points increase in HerQLes score was observed relative to baseline [39].

Permanent synthetic mesh studies demonstrated numerically large and immediate improvements, with an average HerQLes score increase of 42.4 points compared to baseline scores across all time points. While cohorts treated with permanent synthetic mesh experienced early HRQoL improvements (i.e., within the first 12 months), HerQLes scores declined over time in all three studies [4, 46, 50]. For example, in the Zolin 2023 study [50], by the 60 months, the mean difference in HerQLes scores decreased from 48 to 38. Similarly, Rosen 2022 [4] and Miller 2022 [46] reported substantial early improvements (45.01 and 39.1 points increase in HerQLes compared to baseline at 12 months, respectively), which declined to respectively 39.17 and 33.4 points improvements at 24 months.

Overall, considering the longest follow-up in each study, the mean post-operative HerQLes score was 75.25 in Phasix™ Mesh studies and 75.56 in permanent synthetic mesh studies.

The NOS was used to assess the risk of bias among included studies. A summary of these assessments is provided in Supplementary Material 2. The methodological quality of included studies varied between mesh types.

Studies evaluating Phasix™ Mesh generally demonstrated moderate methodological quality, with total quality scores ranging from 5 to 7 out of 9 and a corresponding low-to-medium risk of bias. Among these, Messa 2019 achieved a score of 7 and low risk of bias, with strengths in exposure ascertainment, outcome assessment, and follow-up procedures, although it lacked a well-selected non-exposed cohort and did not adjust for secondary risk factors [48]. The two Christopher studies and the Talwar study shared similar limitations, including the absence of adjustments for key confounders and weaknesses in cohort selection, resulting in medium risk of bias and slightly lower scores [36, 39, 49]. On the other hand, the permanent synthetic mesh studies were generally of high quality, with all three studies receiving low risk of bias ratings. Rosen achieved a score of 9, indicating robust methodology across all domains, including comprehensive adjustment for confounders [4]. Miller and Zolin also performed well, each scoring 7, though they shared some limitations in non-exposed cohort selection and adjustment for secondary risk factors [46, 50].