Abstract
Keywords
Introduction
Traumatic tympanic membrane (TM) perforations are common complaints in otolaryngology clinics. Although spontaneous healing is anticipated in most of the cases, some perforations may take a longer healing time or fail to heal. Persistent TM perforation reduces the quality of life when patients are requested to limit water activity and possibly related sports. Moreover, it carries a risk for the development of chronic infections, acquired cholesteatoma, and facial nerve paralysis.1,2 To avoid long healing time and complications of chronic TM perforations, many studies have evaluated the use of nonsurgical treatments to decrease the closure time including Gelfoam, growth factor, and ofloxacin. It is believed that the application of antibiotic ear drops will facilitate otitis media and the traditional treatment is to keep the ear dry. However, it has been concluded by Lou et al that a moist environment does not contribute to middle ear infections, in fact it promotes the healing and shortens healing time. 3 A broad-spectrum quinolone topical ear antibiotic known as ofloxacin 0.3% (w/v) is used to treat acute and chronic infections of the external auditory canal and middle ear. Animal studies have shown that the use of ofloxacin ear drops accelerated the healing of TM perforation. In addition, ofloxacin has acidic pH and a recent animal study has shown that the use of acidic pH solutions may enhance TM healing.4-6 However, recently there are more studies that evaluated the use of ofloxacin in traumatic TM perforation in humans. Ofloxacin has a low medical cost with a mean of US $15.53 when used in the healing of traumatic eardrum perforation. 7 Therefore, the aim of this comprehensive meta-analysis and systemic review is to assess the effectiveness of ofloxacin drops in traumatic TM perforations in decreasing healing time and its effect on hearing gain.
Methods
This study was performed in accordance with the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA). 8 A study protocol involving a PICO question (P = Population, I = Intervention, C = Comparison, O = Outcome) 9 was developed and registered in the International Prospective Register of Systematic Reviews (PROSPERO) with the registration number: CRD42022377909 to answer the question: In patients with traumatic TM perforations, does the use of ofloxacin ear droplets provide faster healing in comparison to spontaneous healing?
Eligibility
Studies that met the following inclusion criteria were included in this review: written in English language and performed on humans were included. We have included only controlled trials in which patients with traumatic TM perforations regardless of the factor causing the perforation who were treated by ofloxacin droplets and compared to spontaneous healing were included in this review. Exclusion criteria included: (1) Patient with severe otologic disease such as chronic suppurative otitis media or ossicular disruption or patients with craniocerebral injury; (2) studies with no pretreatment values or single-arm clinical studies; (3) articles with foreign language or other study design.
Database Search
Computerized search of 4 online databases was performed (Medline/PubMed, CENTRAL, Clinical Trials.Gov, and Google Scholar) to retrieve all relevant articles from the inception of each database to January 2024. No restriction on language, age, or study design was made. Combinations of the following keywords were used “Ofloxacin,” “fluroquinolones,” and “Tympanic Membrane.” The literature search was carried out by one author (A.H.K.) and revised by another author (M.M.).
Selection of Studies
After removal of duplicates using Endnote 20 (Clarivate, 2013), 2 pairs of reviewers [(M.M. and A.H.K.) and (A.A.A.T. and F.N.)] independently screened the remaining studies’ titles and abstracts for relevant articles. Full-text studies were retrieved and evaluated in pairs for eligibility. In the event of any disagreement between the reviewers regarding the inclusion of a study, a discussion was held, and the opinion of the senior author (H.A.A.) was used to make the final decision. A total of 6 studies were selected for further analysis from the 29 studies through screening the abstract and full texts (Figure 1).
Flow diagram of the search process and search outcomes.
Quality Assessment
Two pairs of authors [(M.M. and A.H.K.) and (A.A.A.T. and F.N.)] independently examined each included study using the Cochrane Risk of Bias Assessment Tool 2 (ROB-2) through Review Manager. Any disagreement among reviewers was resolved by discussion. 9
Data Extraction
Two pairs of authors [(M.M. and A.H.K.) and (A.A.A.T. and F.N.)] independently extracted details from each included trial using a standardized data extraction sheet. Collected data included at least: country, year of publication, number of participants in each arm. Patient characteristics such as age, sex, intervention details, laterality, cause of TM perforation, position of perforation, hearing gain, rate of TM closure, and time for closure. One author (M.M.) entered data, and 2 authors reviewed data entry for accuracy (A.H.K. and F.N.).
Outcome Measures
The primary outcome measures were the closure rate, closure time, and hearing gain in the ofloxacin treatment group versus the closure rate, closure time, and hearing gain. The secondary outcome was the infection rate.
Statistical Analysis
We summarized variables with a normal distribution as mean [standard deviation (SD)] and variables with skewed distribution with medians [interquartile range (IQR)]. Categorical variables were presented as numbers (percentage). Whenever necessary, we imputed means and standard deviations from reported statistics using established methods.10,11
We used a random-effects model to summarize estimates across studies, accounting explicitly for the between-study heterogeneity. 12 For the main analysis, we used an inverse variance model with the between-study variance estimated via the restricted maximum-likelihood method. 13 As a sensitivity analysis, we calculated summary estimates under an inverse-variance fixed-effect model. 12
Whenever feasible, treatment effects were summarized via the mean difference (MD) with 95% confidence intervals (95% CIs). When trials reported the same outcome using different scales, the standardized mean difference (SMD) was used instead. SMDs were converted back to a more clinically interpretable scale using an estimate of the population standard deviation. 14 The latter estimate was obtained by pooling the sample standard deviation from control groups from all eligible studies. For binary outcomes, the relative risk (RR) was used as a metric due to its more straightforward clinical interpretation. We computed the number necessary to treat (NNT) when appropriate. 15
We tested the presence of statistical heterogeneity among study estimates via Cochran’s Q test 16 and quantified the between-study variance with the I2 statistic. 17 A P value <.10 was considered statistically significant heterogeneity for the Q test. 15 To facilitate the interpretation of the heterogeneity, we used 95% predictive intervals (95% PI), which describe a range with a 95% probability of containing the future treatment effects to be estimated in a new study or setting, considering the observed heterogeneity and the uncertainty in the current study estimates. 18
Since the number of examined studies was less than 10, we did not conduct assessments of small-study bias and publication bias as previously recommnended. 19
All analyses were conducted in Stata (version 18; StataCorp LLC). For all summary treatment effects, a P value less than .05 (2-sided) was considered statistically significant.
Results
Characteristics of the Included Studies and Participants
A total of 6 studies (involving 502 participants) contributed data for analysis.3,7,20-23 The median (IQR) number of participants was 86 (47 to 91). Four studies had a follow-up duration of 26 weeks, while 2 had a follow-up of 13 weeks. The mean (SD) age of the participants 34.6 (9.2) years and 305 (61%) females (Table 1).
Characteristics of the Included Trials (n = 6).
Abbreviations: SD, standard deviation; OD, once daily; BID, twice daily; w/v, weight/volume.
Risk of Bias in Included Studies
All 6 studies were deemed to have a high overall risk of bias. The primary reasons for this assessment were inadequate randomization method in 1 study (17%) and unclear randomization methods in the remaining 5 trials (83%). Besides, all studies were categorized as having a high risk of selective reporting, as 5 studies provided no information on trial registrations, and 1 trial was registered retrospectively. Lack of blinding for both patients and clinicians led to a judgment of high risk of performance bias in 5 studies (83%). Four trials were considered to have a low risk of bias in outcome measurement because they employed blinded outcome assessors (Figure 2).
Risk of bias assessment (n = 6 studies). D1 denotes domain 1: bias arising from the randomization process (selection bias). D2 denotes domain 2: bias due to deviations from intended interventions (performance bias) D3 denotes domain 3: bias due to missing outcome data (attrition bias). D4 denotes domain 4: bias in the measurement of the outcome (detection bias) and D5 denotes bias in the selection of the reported results (selective reporting).
Outcomes
The effectiveness of ofloxacin on closure rate
Six studies with 7 independent groups reported data on closure rate. Figure 3 presents the summary treatment effects based on the data of 245 participants assigned to ofloxacin and 257 to the control group (observation). Under a random-effects model, patients treated with ofloxacin were associated with an 18% increase in the chance of closure of perforations compared to those assigned to the observation group (RR: 1.18, 95% CI, 1.08 to 1.28, P < .001). The NNT was 146 (95% CI, 62 to 238). However, there was evidence of statistical heterogeneity across estimates (Cochran’s Q test, P = .05, and I2 = 69.4%), with the 95% prediction interval spanning from 0.94 to 1.48. Results from the fixed-effect analysis led to similar conclusions (RR: 1.13, 95% CI, 1.07 to 1.20, P < .001).
Summary treatment effects of ofloxacin versus observation for closure rate. RR denotes relative risk. 95% CI denotes 95% confidence interval. RE denotes random-effects model. FE denotes fixed-effect model. The 95% predictive interval ranged from 0.94 to 1.48. For Lou et al (2015), treatment effects were extracted separately by groups according to the size of perforations (medium-sized and large-sized perforation groups, respectively).
The effectiveness of ofloxacin on healing time
Closure time was evaluated using the same body of evidence as closure rate (6 trials with 7 subgroups involving 245 participants assigned to ofloxacin and 257 to the observation group). Overall, under a random-effects model, patients treated with ofloxacin were associated with a closure time 21.16 days shorter (95% CI, −28.91 to −13.41, P < .001) than their nontreated (observation) counterparts (Figure 4). The treatments effects exhibited statistical heterogeneity (ie, varied more than expected by chance alone), with a Cochran’s Q test resulting in P < .001 and I2 = 96%. The 95% prediction interval spanned from −49 to +7 days. Results from the sensitivity analysis (fixed-effect) model provided analogous conclusions but with a shorter 95% confidence interval (MD: −18.4 days, 95% CI, −19.96 to −16.82, P < .001).
Summary treatment effects of ofloxacin versus observation for closure time (in days). MD denotes the mean difference in days. 95% CI denotes 95% confidence interval. RE denotes random-effects model. FE denotes fixed-effect model. SD denotes standard deviation. The 95% predictive interval ranged from −49 days to 7 days. For Lou et al (2015), treatment effects were extracted separately by groups according to the size of perforations (medium-sized and large-sized perforation groups, respectively).
The effectiveness of ofloxacin on Hearing gain
Five studies (with 6 subgroups) involving 418 participants (206 in the ofloxacin group and 212 in the observation group) reported data on hearing gain. Under a random-effects model, patients treated with ofloxacin were associated with significantly larger hearing gains over time compared to participants assigned to the observation group (SMD: 0.21, 95% CI, 0.02 to 0.40, P = .03; Figure 5). Assuming a population standard deviation of 4.84 dB, this estimate represents a gain of 1.02 dBs more in the ofloxacin group compared to the observation group (95% CI, 0.1 to 1.94 dB). No evidence of statistical heterogeneity was observed (Cochran’s Q test, P = .28 and I2 = 0%). Results from the fixed-effect model provided virtually identical summary treatment effects (Figure 5).
Summary treatment effects of ofloxacin versus observation for hearing gain (results are in standard deviation units). SMD denotes the standardized mean difference. 95% CI denotes 95% confidence interval. RE denotes random-effects model. FE denotes fixed-effect model. SD denotes standard deviation. The 95% predictive interval ranged from −0.06 to 0.48 (SMD scale). For Lou et al (2015), treatment effects were extracted separately by groups according to the size of perforations (medium-sized and large-sized perforation groups, respectively).
Infection rate
Five studies involving 6 subgroups with 456 participants (220 treated with ofloxacin and 236 assigned to observation) contributed data for infection rates. Patients in the ofloxacin group were associated with a 13% reduction in the risk of infections compared to their observation-assigned counterparts, but this estimate was not statistically significant (RR: 0.87, 95% CI, 0.39 to 1.97). No evidence of statistical heterogeneity was observed, and the fixed-effect model yielded identical results (Figure 6).
Summary treatment effects of ofloxacin versus observation for infection rate. RR denotes relative risk. 95% CI denotes 95% confidence interval. RE denotes random-effects model. FE denotes fixed-effect model. For Lou et al (2015), treatment effects were extracted separately by groups according to the size of perforations (medium-sized and large-sized perforation groups, respectively).
Discussion
This study aimed to summarize the literature to help physicians appreciate possible effects of prescribing ofloxacin ear drops to patients with traumatic TM perforations. No previous studies were conducted in similar fashion. Although the systematic review includes a relatively small number of participants, most of the studied outcomes are powered with significant P values. The closure rate and closure time of traumatic TM perforations were both shown to improve when ofloxacin is added to the treatment regimen. The closure rate was higher in patients receiving ofloxacin ear drops in all included studies. Thus, the pooling of studies for meta-analysis confirmed that ofloxacin has a statistically significant effect on closure rate of TM perforations compared to observation only. Moreover, the healing time was shown to be significantly faster. Based on our study, the MD value for the closure time was −18.4. Although this does not add value to the clinician in short-term perspective, some studies showed an increase in depression and stress in patients with chronic otitis media. 24 Therefore, even small increase in rates of healing with promoting faster healing may improve patients’ satisfaction, earlier return to watersports, and reduce the cost of follow-up. Our analysis of hearing gain was statistically significant and showed an increased in hearing gain by 1.02 dB, which is a change that is not clinically significant. Nevertheless, on complete closure of acute TM perforation, it was reported that there are no significant differences in hearing when compared to hearing tests prior to perforation.1,2
This study had also considered the risk of infection that may result from adding ofloxacin ear drops to TM perforation. Based on our analysis, the ofloxacin group were associated with a 13% reduction in the risk of infections compared to observation group. In the literature, it is shown that increased moisture in ears can cause patient discomfort and result in treatment interruption, 3 and some authors have related excessive ear moisture to cholesteatoma.1,2 However, when ofloxacin is used in appropriate dosing regimen, it provides TM perforation a properly moist environment improving the function of local growth factors and epithelial migration. Ofloxacin offers an acidic environment to enhance the healing of TM perforation while suppressing infection and catabolic activities associated with alkaline environment. 4 The reported application of ofloxacin was 2 drops once daily in 4 studies and twice daily in 2 studies.
Ofloxacin benefits could be summarized as: first, it can be applied readily and is inexpensive. Second, it is a broad-spectrum antibiotic used to treat chronic suppurative otitis media and early grommet insertion otorrhea with no evidence of ototoxicity and systemic side effects.25-27 It is still unknown how ofloxacin quickens the healing process. Since no prior research has demonstrated that ofloxacin by itself speeds up TM perforation closure, we hypothesize that the moist environment provided by ofloxacin helps prevent necrosis and wound edge dehydration and thus speeds up TM perforation healing. A moist wound, according to experience in trauma medicine, prevents necrosis and dehydration at the edges, speeds up the proliferation of granulation tissue, and promotes epithelialization, making the healing process easier.27-29 According to several studies in otology, ear exudate and otorrhea at the edges of perforations help heal traumatic TM perforations,26,30 improve the success rate of myringoplasty,30,31 and prevent necrosis and dehydration. Hyaluronic acid can be applied to the wound to moisten it and speed up TM perforation healing. 32 The moist eardrum environment in the ofloxacin drops treatment group may have contributed to the ofloxacin’s shortened closure time. The correct moisture balance is essential for wound healing. According to a previous study, a balanced moist surface in acute wounds facilitated the action of growth factors, cytokines, and chemokines, promoting cellular growth and the creation of a temporary wound matrix. 33 The migration of keratinocytes into the center of the wound base is encouraged by the increased electrical gradient that a moist wound environment offers between the wound and the wound base and margin. Many biological processes, such as the healthy operation of excitable tissues, left-to-right patterning, organogenesis, and wound healing, depend on electrical gradients.5,20
Limitations
This study had some limitations that require elaboration. On literature review, there were only few properly conducted articles concerned with ofloxacin or similar interventions in healing of traumatic TM perforation that met our inclusion criteria, in which all of them were conducted in China. Therefore, we were able to find only 6 studies and 4 of them were conducted by Lou group. It was not clearly stated in their studies whether these patients were individual or duplicated patients, which may contribute to selection bias. However, the corresponding author was contacted, and it was confirmed that there is no overlap between the studied populations. Furthermore, biases resulting from genetic and ethnic variations could exist. Last, despite recent publication, the sample size is relatively small.
Areas for Further Research
On review of literature, traumatic TM perforations seemingly are not studied sufficiently. This may be related to higher load on clinicians working in trauma centers assessing such cases. However, more research needs to be directed to study the effect of ofloxacin ear drops and sterile saline drops in comparison to observation group to understand whether the moist environment theory contributes to the healing of TM perforation or ofloxacin has a potential role. Moreover, in patients with traumatic TM perforation, the stress and anxiety related to the diagnoses has not been studied and provides a proper field for upcoming research. Also, the cost of such intervention, although briefly mentioned in some texts, is not properly studied bearing the psychological impact of this diagnosis and the benefits such intervention may provide.
Conclusion
Ofloxacin use in patients with traumatic TM perforation is effective in reducing healing time and increasing rate of TM perforation closure. No evidence of increased risk of hearing loss or infection rates are encountered when ofloxacin is prescribed to patients with traumatic TM perforation.
Footnotes
Data Availability
All data generated or analyzed during this study are included in this published article.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Ethical Statement
Ethical approval was not required for this systematic review and meta-analysis. However, this study was performed in accordance with the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA). A study protocol was developed and registered in the International Prospective Register of Systematic Reviews (PROSPERO) with the registration number: CRD42022377909.
Consent to Participate
Informed consent was not required for this systematic review and meta-analysis.
