Audiological and Middle Ear Outcomes in Ukrainian Soldiers with Traumatic Tympanic Membrane Perforation

Introduction

Compared to the past, modern armed conflict differs in terms of form, scale, and complexity [1], and there has been a significant increase in the number of injuries caused by explosives [2,3]. During the U.S. Civil War, only 9% of injuries were due to explosives, with the other 91% due to gunshot. Subsequently, the proportion of gunshot injuries to explosive injuries has decreased, with the latter now predominating. For American soldiers, it has been estimated that in World War I the ratio was 65/35 (gunshot/explosive), in World War II it was 27/73, in Korea 31/69, Vietnam 35/65, and in Iraq and Afghanistan 19/81 [3]. Another study among U.S. servicemen, in which injuries were analyzed by dividing them into blast, gunshot, and other non-combat injuries, blast injuries were 64% in World War II, 54% in the Korean War, and 68% in Iraq [2]. During 2006–2010, 72% of U.S. military compensations were for hearing damage [4]. From Iraq, 31% of veterans were diagnosed with post-explosion ear damage [5]. In 2013, about 551 000 U.S. soldiers suffered blast injuries to the ear from operations in Iraq and Afghanistan [6,7].

The current war in Ukraine is being fought with a relatively large number of explosive devices. The main means of destruction are artillery, guided aircraft bombs, precision-guided aerial and ground drones carrying explosive charges, and missiles. Anti-personnel and anti-tank mines, as well as anti-tank guided missile systems, are most often used for short-range attacks [8]. According to available sources, the Russian army fires up to 60 000 artillery shells per day [9,10], and its airforce drops up to 3000 guided aerial bombs of varying power each month [11].

There are 4 types of blast injuries [12–15]:

Primary blast injuries most often occur to air-filled organs such as the ears and lungs, but can also involve abdominal organs, the parenchyma of the brain, and the eyes. The ears are particularly sensitive to explosions, and injuries to them manifest as hearing loss, tinnitus, dizziness, or bleeding [16]. Common injuries are damage to the ossicular chain, cochlear damage, and perforation of the eardrum (the most common injury to the middle ear) [13,17]. Horrocks (2001) found tympanic membrane rupture in 74% of people diagnosed as having primary blast injury [18]. A generally accepted procedure is to observe the healing of ear injuries for 3–6 months [19]. The physical examination should include direct visualization of the tympanic membrane (otoscopy) and a general assessment of vestibular function and hearing tests [17]. Fortunately, most traumatic tympanic membrane perforations heal spontaneously, in which case the membrane returns to its original function [20]. Spontaneous healing occurs most intensely in the first 3 months after injury and lasts up to 6 months and even up to a year [19–22]. Small perforations are more likely to close completely [23,24]. However, if healing stops, perhaps due to tissue loss, secondary infection, or the formation of pathological lesions, surgical reconstruction of the middle ear is necessary [20,25–29].

The effect of a shock wave of up to 160 dB is likely to cause serious damage to the ear [30,31]. The degree of injury depends on factors such as the speed, intensity, or duration of the pressure pulse [25]. Ear trauma can result in conductive, sensorineural, or mixed hearing loss, indicated in the results of pure-tone audiometry [32].

The most common cause of conductive-type hearing loss is damage to the tympanic membrane. The damage can range from a minor vascular defect, small linear ruptures, to wrapping of membrane fragments into the tympanic cavity; more seriously, there can be large perforations outwards or almost total perforations involving displacement and fracture of the ossicles [30,33]. Mixed or sensorineural hearing losses indicate damage to the inner ear [25,30,31]. Mixed hearing loss occurs in most patients after an explosion, although an audiometric study of U.S. Army veterans found that the majority of those with a hearing loss had sensorineural hearing loss [34].

The size of a tympanic membrane perforation is given in millimeters (mm) or as a percentage of the perforation area to the area of tympanic membrane. In a comparative study of the treatment of post-traumatic tympanic membrane perforation, a criterion based on perforation size was made, such that a small perforation was <3 mm, a medium perforation 3–5 mm, and large perforation >5 mm [35]. The same ranking of perforation sizes was also presented by Jafarov et al [36]. For patients who fail to achieve satisfactory outcomes from tympanoplasty, a final option is the use a hearing implant [37–39].

Therefore, this study evaluated the middle ear and audiological outcomes in 70 men with persistent traumatic tympanic membrane perforation fighting the war in Ukraine.

Material and Methods

ETHICS:

This retrospective study was approved by the Bioethics Committee of the Institute of Physiology and Pathology of Hearing, Warsaw (IFPS: KB/8/2024). All patients voluntarily consented to the processing of personal data under the condition of its protection according to the Law of Ukraine “On Personal Data Protection.” Collected data was anonymized and stored in a dedicated database In Excel.

INCLUSION AND EXCLUSION CRITERIA:

The inclusion criterion was post-explosive ear damage with persistent perforations not exceeding 5 mm (extending up to half the diameter of the eardrum). Exclusion criteria were perforations resulting from chronic inflammatory lesions of the ear or after previous tympanoplasty surgeries, as well as conditions following shrapnel trauma to the temporal bone.

STUDY GROUP:

A retrospective analysis of 70 patients who had persistent post-traumatic ear injury caused by blasts in warfare was performed from July 2024 to March 2025. Patients were aged 19–59 years (M=37.2; SD=10.2). All had perforations remaining after some spontaneous healing had taken place and the lesions had not been operated on before. Details of the sociodemographic and clinical characteristics of the patients are shown in Table 1.

AUDIOLOGICAL EVALUATION:

Otoscopy and pure-tone audiometry (PTA) were used for audiological evaluation. PTA was performed to determine air and bone conduction thresholds. All participants underwent hearing evaluation before and after surgery, but as we present in the following sections of the article, we only analyzed data gathered by us, not from external medical centers, to exclude variability. We obtained 59 preoperative (out of 70 participants) and 19 postoperative tests of pure-tone audiometry. We excluded 2 patients with preoperative deafness from the comparison of hearing test results. After the treatment and convalescence, patients returned to their military units and duties.

SURGICAL TECHNIQUE:

Assessment of perforation size was performed intraoperatively using ear dissectors with an arm length reaching from 2 to 5 mm. Surgeries were carried out under general anesthesia, with a cut made on the posterior wall of the external auditory canal with a Rosen knife. In some patients, it was necessary to extend the bony walls of the ear canal using a diamond cutter. In such cases, the cut was extended to the lower and anterior walls. After lifting the skin and tympanic membrane, it was possible to assess the condition of the middle ear and remove any lesions encountered in the tympanic cavity. Removal of lesions from the middle ear was often associated with an increase in the size of the perforation, especially in cases involving wrapped fragments of the tympanic membrane. These fragments were excised along with part of the edge of the perforation. The next step was to separate the tympanic membrane from the handle of the malleus. If necessary, reconstruction of the ossicles was conducted. Next, a piece of perichondrium was taken from a scrap of the ear. The perichondrium graft was placed under the tympanic membrane ring on the malleus handle under the remnants of the tympanic membrane. The surgical technique involved at least 3 points of support for the eardrum graft: on the handle of the malleus; on the bony walls (posterior, inferior, and sometimes anterior); and under the eardrum ring. A pressure dressing was left in the ear canal for up to 10 days. All surgeries were performed by the same surgeon (the first author).

STATISTICAL ANALYSIS:

Statistical analysis was performed using IBM SPSS Statistics (v. 25) and Jamovi Software (v. 2.6.44). Descriptive statistics were used to indicate the means (M) and standard deviation (SD). The Shapiro-Wilk test was used to assess the normality of the data, followed by a t test, the Mann-Whitney U test, and the Wilcoxon test or Kruskal-Wallis test. P<0.05 indicated a statistically significant result.

Results

RELATIONSHIP BETWEEN THE SIZE OF THE PERFORATION AND PURE-TONE AUDIOMETRY RESULTS:

The size of the tympanic membrane perforation was measured during an intraoperative evaluation. According to the previously mentioned classification [35], perforations were small (<3 mm) in 4 patients (6%) and medium (3–5 mm) in 66 patients (94%). In our analysis (Kruskal-Wallis test) of the effect of the peroration size on hearing results, we discarded subjects who had a 2-mm perforation due to the smaller number of subjects compared to the other perforations. We analyzed the results in relation to ABG (n=56); a difference of 1 mm was not statistically significant (3 mm vs 4 mm – P=0.158 and 4 mm vs 5 mm – P=0.072). For an air-bone gap difference of 2 mm (3 mm vs 5 mm), the result was statistically significant (P=0.002).

RELATIONSHIP BETWEEN PREOPERATIVE PURE-TONE AUDIOMETRY RESULTS AND CHANGE IN AVERAGE THRESHOLD AND AIR-BONE GAP VALUES AFTER OTOSURGERY:

Table 2 shows the mean hearing thresholds and standard deviations for 59 subjects. To assess whether the surgery influenced hearing, a group of 19 subjects was separated and statistically analyzed via the paired-sample t test or Wilcoxon test, depending on the normality of the data. The results were statistically significant. For air conduction, the mean hearing threshold before surgery was worse than after surgery. We noted the shift in air conduction (AC) threshold level compared to bone conduction (BC) threshold level (the closure of the air-bone gap – ABG). As with bone conduction, we noted a change, but we do not see it as clinically significant because the difference was quite small (1.5 dB) and may have been due to the subjective nature of the pure-tone audiometry.

RELATIONSHIP BETWEEN THE NUMBER OF PERFORATED QUADRANTS (1 OR 2 QUADRANTS) AND PURE-TONE AUDIOMETRY RESULTS:

Most patients (42 [60%]) had perforations occurring in 2 quadrants. In the remaining 28 patients (40%), perforations were present in just 1 quadrant.

As illustrated in Table 3, perforation of 2 quadrants compared to 1 quadrant was associated with worse air conduction and larger ABG (n=59); with 2 quadrants, the mean AC was 39.1 dB and the mean ABG was 21.2; while with 1 quadrant, the mean AC was 29.0 dB and the mean ABG was 13.7 dB. The differences were statistically significant, with P=0.004 and P<0.001, respectively (n=59).

RELATIONSHIP BETWEEN THE LOCATION OF THE TYMPANIC MEMBRANE PERFORATION AND AIR-BONE GAP:

Table 4 presents a comparison of values of air-bone gaps in relation to site of perforation. Our study showed that the difference in ABG between anterior (anteroinferior and anterosuperior quadrant) and posterior (posteroinferior and posterosuperior quadrant) perforations were not statistically significant.

PATHOLOGICAL LESIONS: DIVISION INTO GROUPS, LOCATION, QUADRANTS, AND RESULTS OF PURE-TONE AUDIOMETRY:

Tympanic membrane perforations were present in all patients, and in 28 (40%) of them occurred in isolation, that is, without any other additional lesion. This was called the uncomplicated group (UG). However, 42 patients (60%) had additional lesions detected intraoperatively, and these were classified into 2 additional groups. One contained squamous epithelium lesions of the middle ear, and was called the SELG group. It included cystic epidermal lesions (cholesteatoma), epidermal layering, ingrowth from the tympanic membrane onto the walls of the tympanic cavity and ossicles, and wrapping of fragments of the tympanic membrane into the tympanic cavity. This SELG group consisted of 31 people – 22 with cholesteatoma and ingrown epidermis, and 9 without epidermal complications, but with the eardrum fused to the ossicles or walls of the eardrum cavity (which poses a risk of developing cholesteatoma complications in the future). The second group (11 patients) had advanced adhesions of connective tissue and ossicular chain lesions associated with poor conduction, and was called the fibrotic adhesions and ossicle lesions group (FAaOLG).

There were 42 patients (SELG and FAaOLG group) who had lesions and perforations. We detected unorganized epidermal masses or epidermis growing into the tympanic cavity in 17 cases (41%), cystic epidermal lesions (cholesteatoma) in 10 cases (24%), wrapping of the edges of the tympanic membrane into the tympanic cavity in 24 cases (57%), advanced adhesions of connective tissue in 28 cases (67%), and damage to the ossicular chain in 10 cases (24%).

In 13 patients (31%), the above lesions occurred in isolation. However, 17 others (41%) had 2 lesions, and 8 (19%) had 3 lesions. The presence of 4 or 5 lesions occurred in only 2 people (5%). Details of the location of the lesions are listed in Table 5.

Table 6 shows the location of tympanic membrane perforations, which occurred singly (1 quadrant) or in 2 quadrants. Tympanic membrane damage associated with middle ear epidermal lesions occurred in 31 patients, in whom 51 lesions were detected, including 16 (31%) in ears with tympanic membrane perforation in 1 quadrant and 35 (69%) in 2 quadrants.

In ears with identified lesions, perforation was most common in the lower quadrants of AI + PI (cholesteatoma, 6 of 10; ingrown epidermis, 7 of 17; ingrown eardrum edges, 11 of 24; advanced adhesions, 10 of 28). Examples of complications are illustrated in the Figures 1 and 2.

Perforation of the tympanic membrane in ears without additional lesions (n=28) was most common in the anterior and inferior quadrants, occupying the 2 quadrants AS + AI (10 times), followed by AI + PI (7 times). For 1 quadrant, the sum of anterior quadrants was higher (n=7; AS, 3 subjects; AI, 4 subjects) than the sum of posterior quadrants (n=4; PS, 0 subjects; PI, 4 subjects).

Among the 31 patients who had epidermal lesions (SELG), there were 12 (39%) who had epidermal complications not combined with “benign” lesions, of whom 4 (33%) had 1 complication, 7 (58%) had 2 types of complications, and 1 (8%) had all types – cholesteatoma, ingrown epidermis, and wrapping of the edges of the tympanic membrane. The remaining 19 individuals (61%) had connective tissue changes and ossicular defects as well as these complications. Among these 19, 11 (58%) had 1 epidermal complication, 5 (26%) had 2 complications, and 3 (16%) had all 3 complications. FAaOLG consisted of advanced adhesions changes, which appeared in 8 people (73%), 1 person (9%) had ossicular defects, and 2 people (18%) had adhesion lesions as well as ossicular defects.

Fragments of the tympanic membrane most often wrapped around the malleus (21 of 24 cases), and cholesteatoma and ingrown epidermis most often formed around the malleus handle (7 of 10 and 12 of 17, respectively). Figure 2 shows the post-explosive perforation of the tympanic membrane on the first day after injury.

In the SELG group, the average threshold for bone conduction was 21.4 dB, for air conduction it was 39.5 dB, and the mean ABG was 18.1 dB. In the FAaOLG group, these values were 13.4 dB for bone conduction, 33.0 dB for air conduction, and 19.6 for ABG. In the UG group, the values were 13.2 dB for bone conduction, 31.4 dB for air conduction, and 18.2 dB for ABG.

Among the 10 patients diagnosed with ossicular defects, advanced connective tissue lesions (possibly after bleeding into the middle ear cavity) were found in 8 subjects (80%). Epidermal complications were observed in 6 subjects (60%), including 4 subjects (67%) with 1 at-risk lesion each (2 with epidermal changes in the tympanic cavity and 2 with an ingrown tympanic membrane) and 2 (33%) with all 3 epidural complications.

Occurrence of more complications was associated with damage of 2 quadrants, which was statistically significant (P=0.033).

RECONSTRUCTION DETAILS:

The eardrum was reconstructed in 67 patients. In 3 (4%), mobility improved as soon as the ossicular chain was reconstructed. Two types of ossicular reconstruction were performed. In 2 patients with incus dislocation and disconnection with the stapes, ossicular chain reconstruction was performed. A prosthesis was made from the patient’s own ossicle, using the interposition method of the incus, which was prepared and placed on the stapes under the malleus handle. In 1 patient, malleus and incus mobilization was performed in the attic through the gently widened attic wall. The immobilization was probably caused by displacement of the ossicles, ligament tears, and bleeding around them. In 3 patients (4%), the middle ear was inflamed, so the tympanic cavity and ossicles were cleared of inflammation without reconstruction. In 32 ears (46%), we performed intraoperative dilatation of the bony part of the ear canal (canaloplasty) to control the anterior and inferior quadrants of the tympanic membrane. One patient developed a re-perforation following middle ear inflammation in the postoperative period.

Discussion

LIMITATIONS OF THE STUDY:

One of the limitations of the present study is that we could not assess potential confounding factors like pre-existing hearing loss or determine the force of the explosion that led to the rupture of the eardrum. Information about the circumstances of the injury is sensitive. Preoperative HRCT (high-resolution computed tomography) or MRI (magnetic resonance imaging) examinations to evaluate the condition of the tympanic and mastoid cavities were performed according to the surgeon’s decision, so their results were not evaluated in this paper. Also, the results of pathological examination of the removed tissues are not included in our analyses.

Conclusions

Injuries to the middle ear caused by explosions from military operations are frequent and require careful evaluation and observation. Tympanic membrane perforations (the most common combat injury to the middle ear) mostly appear in 2 quadrants, which leads to greater hearing loss and bigger air-bone gap. The difference in ABG was statistically significant when comparing the results of pure-tone audiometry with perforations in 1 quadrant (anterior or posterior) versus 2 quadrants. In contrast, there was no difference in ABG in the anterior and posterior quadrants. Persistent perforations of the tympanic membrane might be accompanied by several pathological lesions, among which the most important are the so-called epidermal complications for the middle ear, which include unorganized epidermal masses or epidermis growing into the tympanic cavity, cystic epidermal lesions (cholesteatoma), and wrapping of the edges of the tympanic membrane into the tympanic cavity. Occurrence of more complications was associated with damage of 2 quadrants. The incidence of complications from blast injuries during warfare may be higher than reported. If complications are suspected, early surgical intervention is recommended. Surgical intervention improves the condition of the middle ear and enhances hearing outcomes, including improved air conduction and closure of the air-bone gap.