It is traditional practice to place a dressing pack within the EAC following myringoplasty, the pack produces a temporary conductive hearing loss until removed. In addition, post-operative ear pack removal is a significant source of anxiety and discomfort, especially in the paediatric age group [2]. However, we have been unable to establish the rationale, or evidence base, for this practice. In fact, there is a paucity of work published on ear packing, except for comparisons of various types of ointment used with the packs. Nevertheless, recent some evidences suggested that the type of EAC packing did not affect the graft outcomes and increase the complications [3, 10].This avoids the discomfort of ear pack removal and an unnecessary out-patient visit, thereby potentially reducing the cost of care. Unfortunately, it is relatively common for surgeons to use antibiotic ointment only in EAC packing, this is believed to prevent the infection of graft [2, 3, 5]. However, the fundamental need for preveting graft infection via use of antibiotic ointment has not been established.

Some scholars recommended that topical antibiotic ointments are effective to use on open wounds in preventing wound infections [11]. However, although the graft infection rate in the NAO group was high than that of UAO group, it was not significantly different among two groups (10.4% vs 4.4%, P = 0.312). Thus, there was no documented evidence of the use of antibiotic ointment affecting graft infection following myringoplasty. This result was similar to previous findings. Hirschmann et al. [12] emphasized that topical antibiotic ointments is ineffective to prevent postoperative wound infections either for dermatologic surgery. Kamath et al. [13] found that there was reduction in the incidence of wound infection with the use of topical antibiotic ointment following surgeries for hip fractures. However, this was not statistically significant to recommend its use in routine practice (relative risk 0.430, 95% confidence interval (CI) 0.120–1.544). Dixon et al. [14] compared the effect of applying ointment to a wound and with no ointment, they found no significant differences in infection rate for all endpoints evaluated (2.3% vs. 1.4%, P = 0.490) and putting ointment on a surgical wound does not benefit the patient. Kalmeijer et al. [15] performed a double-blind, randomized, placebo-controlled study to determine whether use of mupirocin nasal ointment for perioperative eradication of Staphylococcus aureus nasal carriage is effective in preventing the development of surgical site infections (SSIs) and demonstrated that mupirocin nasal ointment did not reduce the SSI rate.

In the field of otologists, Cunningham et al. [16] applied unilaterally coating one grommet with antibiotic ointment to prevent the infection but the contralateral tube was not in 128 children undergoing bilateral tympanostomy tube placement, the incidence of postoperative infection was 4.7%. in untreated grommets and 2.3% in treated side, they believed that the technique of antibiotic ointment application to the tympanostomy tube at the time of surgery was not sufficiently efficacious to prevent infection. Meghji et al. [5] compared postoperation complications rate of the EAC packing using chloramphenicol ointment and packing using bismuth iodoform paraffin paste following tympanoplasty and found both comparable methods. In addition, no use of antibiotic ointment may reduce postoperative symptoms, fullness, and pressure.

In this study, Erythromycin ointment was choosed to pack lateral surface of the graft. Most of literatures reported that Staphylococcus aureus and Pseudomonas aeruginosa were the commonest bacteria for chronic suppurative otitis media (CSOM) [17,18,19,20,21,22,23,24], however, these most common bacteria showed the conflict results for susceptibility and resistance to various antibiotics [18,19,20,21,22,23,24]. Some studies found that Staphylococcus aureus showed high sensitivity to erythromycin [18,19,20] but only a few reported poor susceptibility [21, 22]. Erythromycin ointment was more useful in our department and in otologic clinc in other hospital [23, 24]. Unfortunately, bacterial culture and antimicrobial susceptibility testing were not performed for the patients with purulence discharge in both group in this study.

Although some scholars reported that the graft success rates following myringoplasty was not affected by the presence of preoperative otorrhea [17, 25], postoperative infection increase the flap and free graft necrosis [11, 26, 27]. In this study, the graft became partial or total necrosis in 9/10 patients with middle ear infection. The postoperative infection may affect the graft neovascularization, thereby result in the graft necrosis and absorption, increase the failure rate of graft. Previous study showed that the graft success rate depends principally on the absence of postoperative infection, placement of an adequately sized graft, no graft movement postoperatively, and the absence of any unrecognized middle ear mucosal disease [3]. In this study, the graft success rate was similar among two groups (92.6% vs. 91.0% (61/67). Of the 11 patients with graft failure, only 2 patients were attributed to the technique factor or the insufficient cartilage graft, while 9 were largely due to postoperative middle ear infection with a secondary partial or total necrosis of cartilage graft, thereby resulted in the residual perforation. The results agrees with previous studies. Yamamoto et al. [28] in the histologic study of homograft cartilages implanted in the middle ear found that there was partial absorption of cartilage when inflammatory changes occurred in the middle ear. Mintz et al. [29] observed that perichondritis and cartilage necrosis may result from infection introduced by repeated laryngeal manipulation. In addition, Elwany et al. [26] in the histochemical study of cartilage autografts in tympanoplasty found that the middle ear infection had a strong lethal effect on chondrocytes, and that cartilage grafts survive better in dry ears. They found that that partial or total necrosis of the graft was associated most commonly with Pseudomonas Aeruginosa, B.Hemolytic Stryptococci and Staphylococcus Aureus [26]. However, Staphylococcus Aureus has identified α-toxin as the primary agent causing the rapid death of cartilage cells (chondrocytes) [27]. Fortunately, in the NAO group, mild purulence discharge was observed in one patient at postoperative 3 days and treated with intravenous antibiotic therapy treatment, the graft subsequently survived and the perforation was successfully closed. Thus, it is crucial to early monitor the graft change and early application of antibiotic therapy treatment with mild purulence discharge for graft success.Previous study believed that clinica incorporated grafts had a continued locus minoris resistentiae, decisive therapeutic steps should be taken when the first indication of an inflammation occurs [30]. However, lateral packing or use of antibiotic ointments of graft affect the observation of postoperative early infection and graft change. Thus, technique of antibiotic ointment application to the lateral packing of graft following myringoplasty for preventing postoperative infections and improving the graft success was not sufficiently efficacious to justify routine use.

Apparently, mean ABG gain was not significantly different regardless of the use of antibiotic ointment or not. Also, similar with previous study [3, 10], no graft lateralization, significant blunting or medialization were found in any group during followup period. Although granular myringitis was seen in both groups, it could be well controlled by ofloxacin ear drops plus dexamethasone. In a word, there was no documented evidence of the no use of antibiotic ointment in the EAC affecting graft lateralization, significant blunting, and graft infection following myringoplasty.

These findings may challenge commonly held surgical beliefs about the need for lateral packing following myringoplasty. The advantages of this study is that no patient was lost to follow-up from each group, thus supporting the results. In addition, the graft change may be clearly observed at different followup points by endoscope because of the absence of EAC packing of biodegradable synthetic polyurethane foam. However, the limitations of this study was small sample size, possibility of bias due to lack of blinding, and short-term followup. In addition, bacterial culture and antimicrobial susceptibility testing were not performed for the patients with purulence discharge in both groups.