Ear tubes in children with otitis media reduce parental sick leave from work and socioeconomic costs

Otitis media (OM) is common in small children [1]. OM is roughly divided into OM with effusion (OME) and acute OM (AOM). OME is defined as the presence of liquid in the middle ear without associated signs of ear infection. In AOM, infected middle ear fluid often results in fever, irritability and ear secretion [2]. Among Danish children, 60% have suffered from OM before reaching school age [1].

OM is often self-limiting, but in some cases recurrent AOM (RAOM) (≥ 3 episodes in six months or ≥ 4 episodes in 12 months) [2] or persistent OME (symptom duration ≥ 3 months) may cause disrupted sleep, recurring fever episodes, ear secretion, hearing loss and delayed speech development. These symptoms negatively affect children’s and caregivers’ quality of life (QoL) and increase parental sick leave (PSL) [3, 4].

In Denmark, 16% of all registered PSL was caused by child illness, most commonly infectious diseases [5]. One AOM episode is associated with a mean increase of 1.35 sick leave days for mothers and 0.5 for fathers (p < 0.01) compared with controls. This indicates that OM has socio-economic consequences [6].

Non-surgical management of OM mainly consists of analgesics and antibiotics [7]. However, the 2015 Danish National Clinical Guideline recommends tympanostomy tube (TT) insertion for children aged 0-5 years with persistent bilateral OME and hearing loss with or without delayed speech development and/or RAOM [2]. TT insertion is the most frequent surgical intervention in the Western world [8]. In 2019 (before the COVID-19 pandemic), 28,393 TT insertions were performed in Danish children aged 0-12 years [9]. The majority of these TT insertions were performed in private ear, nose and throat (ENT) specialist clinics.

TT insertion in children with OM is associated with significant parent-observed symptom regression [3] and improved caregiver QoL measured by number of interrupted nights, absenteeism, cancelled social activities and doctor visits [4].

However, PSL has not previously been used as an outcome to measure the effect of TT insertion in children with OM, nor has the financial impact of TT insertion been explored. Thus, the aim of this study was to 1) investigate changes in PSL after TT insertion in children with OM under the age of 12 years; and 2) estimate the overall cost reduction in case of reduced caregiver sick leave (if any) one year following TT insertion.

METHODS

This prospective, observational, multicentre study based on electronic patient-reported outcome (ePRO) data retrieved from the Danish ENT Specialist Organisation database on ear-related symptoms, symptom duration, number of AOM episodes, treatment effect and PSL before and after TT insertion using the Danish National Tympanostomy Tube Insertions Questionnaires (DANTIQ) [10]. A total of 17 Danish private ENT clinics, comprising 26 ENT specialists, assisted with patient recruitment and data collection.

Parents of children < 12 years who were scheduled for TT insertion were offered inclusion. The exclusion criteria were children ≥ 12 years; for parents, the exclusion criteria were no email address, insufficient language abilities, reduced cognitive capabilities or illiteracy and unwillingness to participate.

Data were collected: two days prior to TT insertion (baseline) and 1, 3, 6, 9, 12, 15, 18, 21 and 24 months after TT insertion. Questionnaires were emailed to parents who completed them at home without interference from physicians or clinic staff.

The pre- and post-operative DANTIQ question in focus for this study was: “Did you have any sick leave days because of your child’s ear problems in the past four weeks?”. Possible answers were “Yes”, “No” and “Do not recall”. Multiple answers were not allowed.

Cost estimates

Cost of illness was defined as resource use including both direct costs, e.g., health sector costs, and indirect costs, e.g., decreased or lost productivity [11].

Direct costs

In this study, direct costs comprised general practitioner (GP), private anaesthesiologist (PA) and ENT specialist consultation costs along with treatment and follow-up expenses. GP and ENT specialist consultation frequency was determined by a “minimum necessary” principle, and the costs were calculated based on the Danish official GP, PA and ENT specialist performance price lists from 2021 [12-14]. Estimated costs of antibiotics were based on the daily dosage price of a frequently administered oral antibiotic (amoxicillin), the average weight of a 22-month-old Danish child [15] and assuming a seven-day treatment period.

Indirect costs

The Danish Medicines Council’s Unit Cost Valuation Catalogue was used for estimating the indirect costs of PSL [16]. The average hourly wage for a Danish employee in 2020 was 179 DKK/hour after tax, and a full-time workday was 7.4 hours [13]. Thus, the indirect cost of one sick leave day was estimated to 1,324.6 DKK. To estimate the overall cost reduction achieved by the decrease in PSL from work following TT insertion in a symptomatic child with OM, direct and indirect costs of two potential treatment regimes, A and B, were compared for a 12-month period. A was non-surgical and antibiotics only. This treatment regime is hypothetical, as caregivers and healthcare professionals find it unethical to adopt this regime. B comprised TT insertion and possibly antibiotics, which represents the standard procedure applied today.

Statistics

The Biostatistical Advisory Service at Aarhus University Hospital validated all data calculations and conducted all statistical analyses.

95% confidence intervals (CI) were calculated and a two-sided p < 0.05 was considered statistically significant. The pre- and postoperative odds ratio (OR) for PSL was calculated and adjusted for patient re-entries in the dataset.

Trial registration: not relevant.

RESULTS

Between March 2017 and September 2020, 4,708 children < 12 years were registered in the database. During the study period, 309 children (5.6%) needed TT re-insertion. Parents of 3,918 children (70.5%) completed the pre-operative questionnaire. Post-operative response rates decreased to 19.1% during the 24-month follow-up period [5]. A non-responder analysis showed no significant difference in gender composition. The age of non-participants was significantly higher (× 1.06 (95% CI: 1.007-1.109)) than that of participants [3]. Baseline characteristics are shown in Table 1.

Responses to the pre- and post-operative DANTIQ question regarding PSL are shown in Table 2. OR of PSL after TT insertion was calculated and compared to baseline. PSL risk decreased significantly (p < 0.005) one month after TT insertion compared with baseline; OR = 0.21 (95% CI: 0.19-0.24). PSL risk decreased further during the 24-month follow-up period.

Table 3 compares direct and indirect costs (DKK) for a child with symptomatic RAOM and/or persistent OME undergoing two different treatment regimens: Treatment A (antibiotics only) and Treatment B (TT insertions and possibly antibiotics) within a 12-month period.

Systemic antibiotics were given at least once to 35.2% of the children in the three-month period leading up to TT insertion and 10.9% had antibiotics administered at least once in the month after their TT insertion.

In Treatment A, Amoxicillin was used to calculate the “price per unit”: 40.61 DKK × 0.352 = 14.29 DKK/child/three-month period.

Local antibiotic treatments (such as ciprofloxacin and dexamethasone combination ear drops) are recommended for children after TT insertion. Thus, post-operative antibiotic treatment costs were calculated as follows: 71.9 DKK × 0.109 = 7.83 DKK/child/one month.

As the response option for the DANTIQ question regarding PSL in a four-week period was trichotomic (Yes/No/Do not recall), exact numbers of PSL days were not reported. Therefore, estimates of total pre- and post-operative PSL days were calculated based on the conservative assumption that a “Yes” response was equivalent to one PSL day/child/four weeks.

The preoperative PSL probability was 47.8% (95% CI: 46.2-49.3%). The preoperative PSL probability was assumed to be independent from month to month prior to TT insertion. To estimate the number of preoperative PSL days/child/year, the preoperative PSL probability per four weeks was multiplied by 13 (13 four-week cycles in a year) equal to 6.2 PSL days/child/year.

The post-operative PSL probabilities at 1, 3, 6, 9 and 12 months were 20.5% (19.2-21.8%), 12.7% (11.6-13.9%), 10.7% (9.7-11.8%), 12.8% (11.5-14.0%) and 11.2% (10.0-12.5%), respectively. To calculate an estimate of the number of post-operative PSL days/child/year, the post-operative PSL probabilities per four weeks were calculated for each follow-up period 12 months after TT insertion and thereafter added together: 0.205 + 2 × 0.127 + 3 × 0.107 + 3 × 0.128 + 3 × 0.112 = 1.5 PSL days/child/year.

The total cost estimates for a 12-month period were 9,282.12 DKK for Treatment A (antibiotics only) and 6,163.78 DKK for Treatment B (TT insertion and antibiotics if needed). This corresponded to a difference in costs between Treatment A and B of 3,118.34 DKK/child/year.

DISCUSSION

The primary aim of this study was to investigate the effect of TT insertion in children with OM < 12 years on PSL. The data revealed a significant PSL reduction already one month after TT insertion.

No previous studies have applied change in post-operative PSL as an outcome to measure the effect of TT insertion in symptomatic children with OM. However, an association between OM in children and increased PSL from work was described in previous studies [17, 18]. Barber et al. [18] found that 50% of 2,867 parents experienced more than one medically diagnosed AOM episode in their child during a six-month period. Among those, 73% had been absent from work or had rearranged their working hours. Thus, PSL appears to be a relevant outcome with which to measure the effect of TT insertion in children with OM.

PSL within four-week periods was reported in this study. However, the exact number of PSL days before and after TT insertion was not registered. Therefore, a PSL reduction estimate was calculated based on the decline in the number of “Yes” answers after TT insertion (Table 2). A “Yes” answer was interpreted as being equivalent to only one PSL day every four weeks even though the real number of sick leave days was probably higher. Rare cases of re-tubulation were not included as an expense in Treatment B. However, we believe that this possible expense is counterbalanced by the conservative assumption that a “Yes” answer was equivalent to only one PSL day every four weeks.

Although the significant reduction in PSL one month after surgery compared with baseline suggests a real effect of TT insertion, children with remaining ear problems may be overrepresented in the group of responding parents. Also, post-operative reduction in PSL after TT insertion may be a result of the self-limiting nature of OM. Furthermore, as parents were predisposed to the idea of surgical intervention, referral bias must be considered.

To provide an overall cost-reduction estimate of the decrease in PSL after TT insertion, we compared two hypothetical scenarios. The total cost reduction between the non-TT insertion treatment regime and the TT insertion treatment regime was 3,118.34 DKK in a 12-month period. Decrease in PSL was the main contributor to this reduction. As 28,393 children had TT inserted in Denmark in 2019 [9], TT insertions may have prevented a substantial productivity loss in society (indirect costs) in 2019 alone.

In a previous study, parents to AOM children reported that the condition resulted in substantial use of medical services and loss of workdays [17]. The total costs of one episode of AOM in children from seven European countries was estimated [19]. The direct costs included number of physician and emergency room visits, telephone consultations, hospitalisations and prescription and over-the-counter drug use. Diagnostic tests and surgical interventions were not included. The indirect costs included workdays lost and travel costs. The total costs per AOM episode ranged from 332.00 € (2,470.77 DKK) in the Netherlands to 752.49 € (5,600.23 DKK) in the UK during a 12-month period with indirect costs representing 61-83% of the total episode. Interestingly, the Netherlands, where the lowest costs were recorded, are also known to have one of the highest rates of TT insertions compared with other Western countries [20].

The total cost estimates in this study are rough estimates. Although the post-operative treatment course after TT insertion is well known and understood, the non-TT insertion scenario is hypothetical as most Danish symptomatic children with OM receive TTs if assessed by an ENT specialist. Also, TTs may take longer than 12 months to fall out or may need to be re-inserted, and some children may develop complications to the procedure that require additional and more expensive treatments. These situations will increase the cost of the post-operative treatment course. Consequently, the direct cost calculations are approximations based on a mix of parent-provided ePRO data, ENT specialist knowledge, experience and guesswork. Also, the indirect cost calculation is based on an average hourly wage for a Danish employee.

However, the significant PSL reduction shortly after TT insertion underpins the results of our analysis. The study methodology was thoroughly revised by a group of four experienced private ENT specialists to ensure alignment between reality and the two treatment regimens (Table 3). Furthermore, revision by a professional health economist ensured that proper indirect cost measurements were applied.

CONCLUSION

TT insertion is associated with a significant decrease in PSL 1-24 months after surgery. Overall cost estimates show that the socioeconomic burden of children with OM who receive TTs is smaller than that of children with OM who hypothetically do not. This indicates that TT insertion improves the clinical symptoms of OM in children while contributing to substantial socioeconomic savings.