Introduction
Continuity of care is important in the management of patients with chronic otitis media, particularly those with cholesteatoma. It is recognised that untreated or inadequately treated cholesteatoma may cause erosion of structures within the temporal bone, and lead to hearing loss, dizziness, facial nerve weakness and temporal lobe abscesses.Reference Jung, Alper, Hellstrom, Hunter, Casselbrant and Groth1 Furthermore, non-cholesteatomatous chronic otitis media may cause many of these complications, notably hearing loss.Reference Browning2 It is for this reason that when patients with tympanic membrane retraction pockets are managed conservatively, they are seen on a regular basis. Patients with unilateral chronic otitis media may be at risk of developing these conditions in the contralateral ear,Reference Tsai, Lien and Guo3 and it is possible that early intervention may mitigate damage to the middle ear.
There are no established guidelines on the frequency or length of follow up for patients after surgery for cholesteatoma. Factors such as the surgical approach, location, and confidence in disease clearance and underlying eustachian tube function may all play a role in determining the likelihood of residual and recurrent disease, and thus in determining a recommended follow-up strategy.Reference Gristwood and Venables4–Reference Gaillardin, Lescanne, Morinière, Cottier and Robier6 However, it is recognised that a proportion of patients will not attend recommended follow-up appointments, and this has the potential to detrimentally affect their care.
An awareness of the likelihood of a patient not attending follow up is important as it may affect a clinician's approach to management. For example, a clinician may be more likely to propose ‘second look’ surgery rather than serial diffusion-weighted magnetic resonance imaging (MRI) in a patient who has had canal wall preserving tympanomastoidectomy. A surgeon may feel that if a patient is unlikely to attend follow up, then it is more appropriate to perform a canal wall down procedure, and may consider obliteration of the mastoid cavity. If there is doubt whether a retraction pocket should be monitored or operated upon, consideration of the likelihood of maintaining follow up is essential.
We audited patients with chronic otitis media treated at our institution who were lost to follow up. This study group were compared with a control group of patients who completed follow up and were formally discharged from care. The study aimed to determine if there were demographic factors that were associated with a higher rate of loss to follow up.
Materials and methods
A retrospective case–control study was conducted. A database (instituted in March 2009) comprising prospectively collected information on patients who were either imaged for suspected or confirmed cholesteatoma or underwent tympanomastoid surgery was searched for patients who were lost to follow up or formally discharged. A patient was deemed to be lost to follow up if there was no record of them having been seen within a year of the intended date of follow up indicated at the most recent appointment they attended.
Fifty consecutive patients who were lost to follow up were identified. These were matched with a control group of 50 consecutive patients who were formally discharged. Demographic information collected included age, sex and postcode. This was used to calculate deprivation indices in the domains of income, health, education, employment and the broad ‘index of multiple deprivation’.7 These data were obtained from the UK Department of Communities and Local Government.8 These data come in the form of a national ranking for each measure among ‘Lower Layer Super Output Areas’. Whilst these data do not relate directly to an individual, this is a commonly used measure from census data.9
Statistical analysis involved the use of the Mann–Whitney U test to determine the significance of the non-parametric ranked data (deprivation indices). Fisher's exact test was used for the categorical data (patient sex and management). Student's two-tailed t-test (unpaired) was used to determine the significance of the continuous parametric data (patient age).
Results
The mean age of patients in the entire sample was 38.1 years. Forty-two per cent were female and 58 per cent were male. Ninety-three per cent of patients underwent imaging and 83 per cent underwent operative management of chronic otitis media.
A total of 410 patient records were searched. This revealed 50 patients within the database who had been lost to follow up (12.2 per cent); these patients formed the study group. In the study group, 82 per cent underwent operative management and 96 per cent underwent imaging. In the control group, 84 per cent underwent operative management and 90 per cent underwent imaging. These differences were not significant (Fisher's exact test: p = 1 and p = 0.44).
There was a significant association between patient age and loss to follow up. The mean age of patients in the study group was 33.7 years (standard deviation (SD) = 16.5 years) and in the control group was 42.5 years (SD = 18.5 years). There was a statistically significant age difference of 8.8 years (p < 0.02, 95 per cent confidence interval = 1.8–15.8 years). Sixty-two per cent of the study group were male, compared with 54 per cent of the control group; this difference was not significant (p = 0.54).
Deprivation indices were calculated for all patient addresses based on postcodes. Although patients who were lost to follow up lived in areas that were more deprived on average, this factor was not statistically significant. Specifically, there was no significant correlation for the index of multiple deprivation, income, health or employment deprivation. Patients who were lost to follow up were, however, statistically more likely to come from an area with higher education deprivation (p < 0.05).
Discussion
Loss to follow up is a significant problem in the management of chronic otitis media, especially in cholesteatoma management. This is particularly important given the more recent changes to the nature of tympanomastoid surgery – canal wall preserving surgery and diffusion-weighted MRI rely on lengthy follow-up strategies. Our rate of loss to follow up (12.2 per cent) correlates with those of other series from the UK and abroad. For example, in Sajjadi's series from the USA, the rate of loss to follow up was 11.6 per cent.Reference Sajjadi10 However, given that patients currently being treated (i.e. neither discharged nor lost to follow up) are included in the database, our reported rate is likely to represent a significant underestimation of the proportion of patients who will at some stage be lost to follow up. Our institution's rate of loss to follow up is, however, not the focus of this study. Rates of follow up will vary between institutions based on patient factors, and the availability, timing and administration of out-patient clinics. We focused on determining patient factors that may be significantly associated with loss to follow up at our institution.
Administrative factors are likely to play a significant role in the likelihood of a patient being lost to follow up.Reference Nash, Kenway and Mochloulis11 Hospitals try to develop administrative systems over time, to decrease the incidence of patients not attending appointments, and to correlate the supply and demand for appointments. During the study period, our institution trialled a number of techniques to contact patients, and arranged evening and weekend appointments.
There are two notable limitations to consider regarding the data presented. Firstly, the use of postcode data to determine five indices of deprivation means that we only get an indirect measure of this information. Whilst postcode data are widely used for these tasks, the data only yield an average for a group of people who live in a similar area, rather than patient-specific information. The catchment area for our centre includes wealthy and deprived areas, and is therefore an appropriate place to study this association.
The second principal limitation is that it is difficult to determine whether there was a significant difference between the study and control groups in terms of disease burden, which may affect follow up. We did not study disease factors or operative outcomes in detail because of the heterogeneity of the outcome data. It is reassuring, however, that there was no significant difference in the proportions of patients in each group who underwent imaging or operative management. The study group was skewed towards cholesteatomatous disease, although a small minority of patients were treated for tympanic membrane retractions or perforations. Nevertheless, this means that these data may not be as applicable for patients with mucosal disease.
There are a number of possible reasons why a patient group may be prone to be lost to follow up. Younger patients may be more likely to move (around an area, a country or indeed internationally), and thus may not receive notification or attend appointments.Reference Pal, Taberner, Readman and Jones12 Patients without employment may also be more likely to move to seek employment. Various patient groups may have fewer concerns about their health,Reference Thornton and Ballard13 or indeed a less productive relationship with healthcare providers.Reference Murdock, Rodgers, Lindsay and Tham14 They may be less likely to engage with their general practitioner, or advise their clinicians of a change in address. Exploring these factors is important, but is beyond the scope of this article.
• Regular follow up is often required for chronic otitis media management
• Follow-up strategies depend on the treatment required
• It would be beneficial to identify those patients likely to be lost to follow up
• Younger age and living in an area of education deprivation are associated with loss to follow up in chronic otitis media patients
• This information can help guide decision making
The findings of this article are important to consider when planning and performing tympanomastoid surgery. They mirror anecdotal reports regarding the probability of a patient attending follow-up appointments. Otologists should consider the follow up required for a proposed course of treatment, and then consider whether their patient would be likely to follow the proposals. If a patient is thought to be at higher risk of being lost to follow up, proactive measures to keep in contact with the patient could be instituted.
Conclusion
Younger patient age and living within an area of education deprivation are associated with a decreased likelihood of attending follow up when being investigated and treated for chronic otitis media. Otologists should consider the follow up required for a proposed management plan, and the likelihood of their patient attending that follow up.
