Introduction
Coronary artery bypass grafting has been performed for 40 years. The incidence of severe hearing loss following coronary artery bypass surgery has been estimated as 0.1 per cent.Reference Arenberg, Allen and DeBoer1 The first reported case of such hearing loss was presented by Arenberg et al. in 1972.Reference Arenberg, Allen and DeBoer1 Sudden sensorineural hearing loss (SNHL) is a well recognised phenomenon that has been attributed to a variety of aetiologies. Several aetiologies have been suggested to explain SNHL following open heart surgery, including: thromboembolic phenomena; hypotension or perfusion failure; hypothermia; ototoxic drug use; and related central nervous system injury (such as stroke). However, no proven pathogenesis has yet been established.
The aim of this study was to investigate hearing changes following open cardiac surgery, using high frequency audiometry, and to determine which clinical and operative factors were related to SNHL.
Materials and methods
Between August and September 2004, 30 patients underwent open cardiac surgery, 20 of whom were capable of completing pre- and post-operative questionnaires. These questionnaires specifically enquired about any changes in patients' hearing, and about the presence of tinnitus. Audiometric testing was performed on post-operative days six to 15. Consequently, 40 ears of 20 patients were included in the study. Patients without a past medical history of hearing problems, ototoxic drug use or neurological disease were examined prior to and following their coronary bypass procedure, and audiometric assessment was also performed. Patients with a history of hearing problems, ototoxic drug use or neurological disease, or of post-operative neurological complications, were excluded from the study. Similar anaesthetic agents were used on all patients, and each anaesthetic parameter was examined.
Audiometry
Audiometric assessment was performed within the Ankara Numune Education and Research Hospital otolaryngology department, pre- and post-operatively. Pure tone audiometry findings were obtained for the frequencies of 500, 1000, 2000, 4000 and 8000 Hz, and also for the following high frequencies: 10 000, 12 000 and 14 000 Hz. All audiometry was performed by the same person using an Interacoustic AC 40 audiometer (İnteracoustic Company, Assens, Denmark), which was calibrated periodically.
Audiometric measurements were compared both pre- and post-operatively, taking into account the following patient parameters: age; sex; presence or absence of diabetes mellitus, hypertension and high cholesterol level; history of myocardial infarction; and presence of coronary artery or valve disease. When comparing pre- and post-operative audiometric results, the following surgical parameters were also noted: type of procedure; cardiopulmonary bypass time; cross-clamping time; and operation temperature (patient's rectal temperature).
Statistics
Data were analysed using non-parametric statistics. Paired comparisons within a group were analysed with the Wilcoxon signed rank test. A p value of <0.05 (two-tailed) was considered significant. The Statistical Package for the Social Sciences version 11.0 for Windows software (SPSS Inc, Chicago, Illinois, USA) was used for statistical analysis.
Results
Twenty patients completed the study. Twenty-five per cent (five of 20) of the patients were female and 75 per cent (15/20) were male. The mean age was 48.8 years (range, 42 to 55 years). Thirty per cent (six of 20) of the patients had hypertension, 20 per cent (four of 20) had diabetes mellitus, 25 per cent (five of 20) had hypercholesterolaemia, 25 per cent (five of 20) had a history of myocardial infarction, 55 per cent (11/20) had coronary artery disease and 45 per cent (nine of 20) had cardiac valve disease.
The mean duration of total cardiac bypass time was 66.45 minutes (standard deviation (SD) 23.3), the mean cross-clamp time was 40.5 minutes (SD 16.8) and the mean operation temperature (patient's rectal temperature) was 30.3°C. Patients' audiograms were analysed to assess pre- and post-operative differences, for each frequency (Figure 1). The mean pre- and post-operative hearing values, for right and left ears, were also evaluated for each frequency (Figures 2 and 3).
Fig. 1 Distribution of median pre- and post-operative audiometric hearing thresholds, by frequency. Pre = mean pre-operative values for 40 ears; post = mean post-operative values for 40 ears
Fig. 2 Distribution of median hearing thresholds for the right ears, by frequency. Pre = pre-operative values; post = post-operative values
Fig. 3 Distribution of median hearing thresholds for the left ears, by frequency. Pre = pre-operative values; post = post-operative values
For the right ear, differences between pre- and post-operative hearing values were statically significant at 500, 1000, 8000, 10 000, 12 000 and 14 000 Hz frequencies (p < 0.05). For the left ear, differences between pre- and post-operative mean hearing values were statistically significant at 1000, 8000, 10 000 and 12 000 Hz frequencies (p < 0.05).
Risk factors for both the individual patients and the surgical procedures were compared with respect to significant hearing loss in each frequency. The following significant correlations were found: between hearing loss and the presence of hypercholesterolaemia, at 8000 Hz (p = 0.03) and 10 000 Hz (p = 0.02) frequencies; between hearing loss and a history of myocardial infarction, at 10 000 Hz (p = 0.05) and 12 000 Hz (p = 0.04); between hearing loss and the presence of hypertension, at 8000 Hz (p = 0.03) and 10 000 Hz (p = 0.01); and also between an increased cross-clamp time and hearing loss at 500 Hz (p = 0.03). We found no differences between patients undergoing coronary artery bypass surgery and those undergoing cardiac valve replacement.
When total mean pre- and post-operative hearing thresholds of the right and left ears were compared for each separate frequency, significant differences were found at 500, 1000, 8000, 10 000, 12 000 and 14 000 Hz frequencies (p < 0.05) (Figure 1).
Discussion
In the general population, the incidence of sudden SNHL is reported as approximately one in 5000. The main aetiological factors are inflamatory vascular, metabolic, biochemical, traumatic, tympanic membrane rupture and idiopathic causes.Reference Byl2
Sensorineural hearing loss following open cardiac surgery was first described by Arenberg et al. in 1972.Reference Arenberg, Allen and DeBoer1 In 1975, Wright and Saunders reported a similar case.Reference Wright and Saunders3 A large, retrospective case analysis including 7000 patients was carried out by Plasse et al. They reported the incidence of SNHL after cardiopulmonary bypass surgery as being one in 1000.Reference Plasse, Mittleman and Frost4
Shapiro et al. categorised patients suffering hearing loss following cardiac surgery into three groups: those with no change; those with slight changes; and those with average hearing deficits of more than 10 dB. Hearing loss of less than 10 dB was difficult to recognise because of poor cooperation and concentration. However, in Shapiro and colleagues' series of 68 patients, 13.2 per cent showed hearing loss at high frequencies after cardiopulmonary bypass surgery.Reference Shapiro, Purn and Raskin5
The incidence of complications following open cardiac surgery is high; surgical complications constitute 50 per cent, and cerebral complications 2–5 per cent. Hearing loss is observed as a post-operative complication in 0.01 per cent of patients.Reference Young In, Mahta and Lowry6 Ness et al. compared the incidence of SNHL following coronary bypass surgery (0.14 per cent) with that seen in a normal population (0.02 per cent).Reference Ness, Stankiewicz, Kaniff, Pifarre and Allegretti7 These authors postulated that the higher incidence of SNHL seen after cardiac surgery was due to risk factors in some patients (diabetes mellitus, hypertension, hypercholesterolaemia, etc), complications due to surgery and the use of ototoxic drugs.Reference Ness, Stankiewicz, Kaniff, Pifarre and Allegretti7 In our study, patients who suffered significant surgical complications and those using ototoxic drugs were excluded from analysis. Furthermore, when risk factors attributed to patients and to the surgical procedure were analysed separately, some risk factors were thought to be causative for SNHL. There are four main theories explaining SNHL after open cardiac surgery: microemboli; intra-operative hypotension or perfusion abnormality; hypercoagulability; and use of ototoxic drugs.Reference Young In, Mahta and Lowry6–Reference Brownson, Stroed and Carver8
Some animal studies have shown that thrombosis, emboli or spasm affecting the labyrinthine artery may cause SNHL.Reference Alford, Shaver and Rosenberg9, Reference Belal10 This makes temporal bone histological studies mandatory in order to clarify the aetiology of SNHL following cardiac surgery.
• The aim of this study was to investigate changes in hearing, using high frequency audiometry, following open cardiac surgery
• Open heart surgery using cardiopulmonary bypass leads to significant post-operative changes in hearing levels at some frequencies
• The presence of hypertension or hypercholesterolaemia, a history of myocardial infarction, and cross-clamp time were found to significantly influence hearing levels following cardiac surgery
• Further studies are required to investigate the aetiology of hearing loss following cardiac surgery
Phillipps and Thornton, in their study of 20 patients undergoing cardiac surgery, described post-operative SNHL mainly at higher frequencies.Reference Phillipps and Thornton11 Any pathology affecting the cochlea and the distal portion of the VIIIth cranial nerve will cause more of a disturbance to the basal portion of the cochlea. Therefore, high frequency audiometry is an effective tool in demonstrating these pathological changes.Reference Belgin, Böke and Dalgıç12 Our findings also revealed that hearing loss was more frequently present at higher frequencies.
In 1967, Jaffe first described SNHL in 14 patients following non-otological, non-cardiac surgical procedures performed under general anaesthesia.Reference Jaffe13 Similar cases have been reported in the literature.Reference Young In, Mahta and Lowry6 All our patients underwent similar anaesthetic procedures; therefore, the effect of general anaesthesia in itself was minimised.
Conclusion
Our findings demonstrated that SNHL occurs after cardiac surgery, especially at higher frequencies. It is thought that some risk factors may be causative for such SNHL, but a definite aetiology could not be clarified. However, we believe that more extended patient investigations and post-mortem histopathological studies may be able to identify the aetiological factors responsible.
