xRead - January 2023

(13.8%, 95% CI 4.5%–27.1%) was higher than short-term (8.0%, 95% CI 4.2%–13.4%), although this difference was not statistically significant ( P 5 0.854) (Fig. 4a of b).

Ribeiro (2015) 17 4 4 27.5 21.3 – – – – – 0.0 –

Moser (2014) 16 11 13 – – – – – – – – –

Herdman (1990) 5 7 9 37.2 – 10.0 – – – – 22.2 25.0 Katzke (1982) 14 6 6 – – – 100.0 – 100.0 – – – Lin (2005) 15 21 25 28.7 – 12.5 – 56.0 – 84.0 12.0 0.0

Suzukawa (2007) 19 4 4 31.0 2.3 – 100.0 – 100.0 – – – Tos (1979) 20 11 11 – – – – – – 63.6 0.0 0.0 Tos (1986) 8 19 22 – – – 45.5 29.4 81.8 58.8 – –

Becker (1998) 13 47 53 – – – 60.4 31.6 90.6 60.5 5.7 7.9

Ghani (2013) 10 12 14 29.9 12.0 19.3 50.0 33.3 92.9 55.6 0.0 33.3

Slattery (1997) 18 11 14 24.8 8.5 18.4 50.0 12.5 100.0 62.5 0.0 60.0

Restenosis

Long-Term (%)

DISCUSSION Postinflammatory medial canal fibrosis is character ized by formation of a fibrotic plug on the TM that oblit erates the medial EAC, leading to a significant CHL. Surgery remains the preferred treatment for late stage PIMCF; however, long-term outcomes remain unclear. Current available evidence is largely comprised of level IV retrospective case series, many of which combine patients with all subtypes of MCF and fail to report long-term data. To gain improved insight into these issues, we retrospectively reviewed our experience with surgical management of PIMCF and specifically sought to report long-term outcomes. Additionally, the current study presents the first level III evidence in the form of a systematic review and meta-analysis of patients treated surgically for PIMCF. For the institutional cohort and meta-analysis, significant improvement in both short- and long-term ABGs were noted when com pared to preoperative ABGs. Although rates of long-term complete restenosis were relatively low in our cohort and the aggregate analysis, the incidence of partial canal stenosis at long-term follow-up approached 60%. In light of these findings, it is not surprising that aggre gate pooled data demonstrated a significant deteriora tion in hearing from short-term to long-term. Study Characteristics Demographically, our study population is consistent with those presented in other reports in the literature discussing management of PIMCF. Our relatively small sample size (21 ears) speaks to the rarity of this disease. There was close to a 2:1 preponderance of females to males, a trend that is well-documented in patients with PIMCF. 1,7–9 Approximately one-third of our patients were treated for bilateral disease, also a common finding that underscores the underlying systemic inflammatory nature of this disease. Although only one patient had a documented dermatologic condition in our cohort, der matologic disease is a common finding in these patients reported in 0% to 53% of PIMCF patients. 1,16 The systematic review and meta-analysis performed included 11 articles (spanning years 1979–2015), 153 patients, and 175 ears. Importantly, we excluded any article including treatment of general canal stenosis that did not fit the pathology of MCF, 4,21,22 and additionally excluded those articles detailing patients with MCF related to causes other than chronic inflammation/OE/ CSOM. 23–25 All included articles were retrospective case series, with cohorts ranging from four to 53 ears. Audiometric Outcomes Audiometric outcomes in our cohort and the aggre gate analysis were strikingly similar. They support the efficacy of surgery for PIMCF but also highlight the risk of recurrence of hearing loss over time. Postoperative

Restenosis

Short-Term (%)

ABG 20

Long-Term (%)

ABG 20

Short-Term (%)

Postoperative Long-Term † ABG (dB) ABG 10 Short-Term (%) ABG 10

Long-Term (%)

TABLE II.

Mean

Articles Satisfying Inclusion Criteria With Outcomes.

Mean

Postoperative Short-Term* ABG (dB)

Mean

Preoperative ABG (dB)

(N)

Ears

ABG 5 air–bone gap; dB 5 decibel.

Patients (N)

*Short-term < 2 years.

† Long-term > 2 years.

Author (Year)

Laryngoscope 127: February 2017

Keller et al.: Retrospective Review and Meta-Analysis

492