Expensive today but cheaper tomorrow: lifetime costs of an active middle ear implant compared to alternative treatment options

Krohn, Markus et al. Article — Published Version Expensive today but cheaper tomorrow: lifetime costs of an active middle ear implant compared to alternative treatment options The European Journal of Health Economics Provided in Cooperation with: Springer Nature Suggested Citation: Krohn, Markus et al. (2024) : Expensive today but cheaper tomorrow: lifetime costs of an active middle ear implant compared to alternative treatment options, The European Journal of Health Economics, ISSN 1618-7601, Springer, Berlin, Heidelberg, Vol. 26, Iss. 5, pp. 797-816, https://doi.org/10.1007/s10198-024-01743-6 This Version is available at: https://hdl.handle.net/10419/323291 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ Vol.:(0123456789) The European Journal of Health Economics (2025) 26:797–816 https://doi.org/10.1007/s10198-024-01743-6 ORIGINAL PAPER Expensive today butcheaper tomorrow: lifetime costs ofanactive middle ear implant compared toalternative treatment options MarkusKrohn1 · KlaasKiesewetter2· AnnikaBuchholz3· BettinaSchlick2· SusanBusch3· ThomasLenarz3· AnkeLesinski‑Schiedat3· HannesMaier4· CorneliaBatsoulis5· MichaelUrban2· SteffenFlessa1 Received: 30 July 2024 / Accepted: 14 November 2024 / Published online: 6 December 2024 © The Author(s) 2024 Abstract Background When choosing between different treatment options, implants often appear too costly. However, this perspective does not take future costs into account. This article evaluates lifetime costs for different surgical interventions to treat hearing loss. Methods The analysis focused on three groups from the perspective of health insurers. Group 1 comprises patients who have only been implanted with a middle ear implant. Patients in Group 2 had already undergone middle ear surgery to improve hearing prior to the implantation of a middle ear implant. Group 3 consists of patients who were treated exclusively with hearing-improvement surgeries (no implant). The lifetime costs were calculated using the Monte Carlo simulation. The inputs were based on medical data from a maximum-care hospital and data from the German healthcare system. Results Based on an average observation period of 26.73years, the lifetime costs amounted to 28,325€ for group 1, 32,187€ for group 2 and 28,381€ for group 3. While the mean values between groups 1 and 3 appear comparable, group 1 has a significantly lower standard deviation (G1 vs. G3: 6120€ vs. 10,327€). Discussion/conclusion Choosing a treatment option can be a complex medical decision and impose a substantial economic burden for the statutory health insurance. Hence, treatment decisions should be patient-centred at first but also including a shared-decision making on economic feasibility, whether proposed treatment alternatives are likely to be successful and economically reasonable. Keywords Middle ear implant· Cost analysis· Lifetime cost· Monte-Carlo-Simulation· Hearing-improvement surgeries JEL Classification I10· I11· I13· C15 Introduction Several treatment options are usually considered as part of any medical decision-making process. For example, treatment options for progressive hearing loss may include hearing-improvement surgeries (tympanoplasties) or middle ear implants. While typical surgery for hearing loss costs around 3700 € (example DRG D30A) [1], the costs for a middle ear implant are higher at around 13,000 € [2]. At first, the implant appears to be disadvantageous from an economic perspective. However, there are other possible effects that may arise in the further course of treatment; therefore, costs should be calculated over the entire lifespan based on medical data. According to the World Report on Hearing by the World Health Organization (WHO), hearing loss is the fourth * Markus Krohn [email protected] 1 Chair ofGeneral Business Administration andHealth Care Management, University ofGreifswald, Greifswald, Germany 2 MED-EL Medical Electronics, Innsbruck, Austria 3 Department ofOtolaryngology, Hannover Medical School, Hannover, Germany 4 Clinic forLaryngology, Rhinology andOtology, Hannover Medical School, Hannover, Germany 5 MED-EL Medical Electronics, MED-EL Research Center, Hannover, Germany 798 M.Krohn et al. highest cause of disability globally, with an estimated annual cost of more than 980 billion dollars [3]. In Germany, the annual cost of health disorders caused by hearing impairments were estimated to be € 2.65 billion in 2011 [4]. Current data suggest that approximately 5% of the world’s population or 430 million people require auditory rehabilitation to address their disabling hearing loss. It is estimated that by 2050 more than 700 million people – or 1 in every 10 people – will suffer a disabling hearing loss (hearing loss greater than 35 decibels (dB) in the better-hearing ear). According to the meta study “Hearing Loss – Numbers and Costs” published by hear-it AISBL, there are approximately 34.4 million people in the European Union living with a disabling hearing loss, with 22.6 million people remaining untreated. This results in an annual overall cost of 185 billion Euros or 8200 Euros per affected person [5]. Furthermore, approximately 10 million people in the EU suffer a hearing loss of 50dB or greater, with around 5.5 million of these still untreated. In 2021 in Germany alone, approximately 4.5 million people (5.35%) had a disabling hearing loss, according to the most recent data from the Institute for Health Metrics and Evaluation (IHME) [6]. In 2019, Löhler etal. investigated six studies involving hearing disorders in Germany, showing a broad range of hearing loss prevalence (between 16 and 25%) with no standardized definition of hearing loss [7]. Also in 2019, Schmucker etal. conducted a systematic review to complement the results of Löhler etal. reporting a prevalence of hearing loss in the child and adolescent German population ranging from 0.1 to 128 per 1000 children and adolescents [8]. Self-reported hearing loss at a higher threshold (> 30dB HL) ranged between 2.4 and 5.2% [8]. Another epidemiological study on hearing status in northwest Germany (HÖRSTAT) estimated the prevalence of hearing loss, as defined by the World Health Organization, to be 16%, which is in line with the number of participants who had a hearing level greater than 40dB in the “hearit AISBL” article [9]. It is obvious that disabling hearing loss negatively affects multiple aspects of an individual’s life when left unaddressed or when the individual’s communication needs are unsupported. For many years, standard treatment for conductive and/ or mixed hearing loss has consisted of middle ear surgeries with or without the replacement of defective ossicles (tympanoplasty) and if possible, the application of a hearing aid. Middle ear surgeries, such as a tympanoplasties and/or mastoidectomies (surgeries to the mastoid bone for infection and the tympanic membrane), are frequently performed with excellent results for the treatment of infectious middle ear diseases with or without implanting passive ossicular prostheses to restore hearing. Middle ear surgeries are primarily concerned with resolving the discharging pathology, as in the case of patients with chronic otitis media (COM), or with the complete eradication in case of cholesteatomas. As the pathologies have a high chance of reoccurrence, these procedures may usually require repeated surgeries. Success of middle ear surgeries can be linked to the functional restoration of the tympanic membrane, with high recurrence rates between 50 and 100%, including a variety of different and frequently interrelated causes of failures, making surgical prognosis very difficult, often resulting in a severe mixed hearing impairment in patients with COM often not able to wear their hearing-aids. In addition, hearing aid usage in such patients may lead to radical cavity infections or recurrent myringitis if sealed too tightly [10]. Even when patients undergo multiple appropriate middle ear surgeries, hearing improvement sometimes remains inadequate. These patients with unsuccessful conventional hearing rehabilitation and a variety of middle ear conditions can be treated effectively using an electromagnetically active middle ear implant such as the Vibrant Soundbridge (VSB), as the anatomical conditions in such cases often impede an adequate acoustic coupling. Active middle ear implants (AMEIs) are medical devices intended to treat hearing loss by stimulation of the middle and inner ear structures. This study investigated the use of the Vibrant Soundbridge (VSB) (MED-EL Austria), the world’s, up to now, most utilized AMEI. With more than 25years of experience in restoring hearing for thousands of people around the world, the VSB is suitable for different types of hearing loss and is also approved for children. It is implanted in cases of sensorineural hearing loss when the fitting of conventional hearing aids is not possible because of chronic otitis externa, as well as in cases of conductive or combined hearing impairment when conventional tympanoplasties and passive ossiculoplasties cannot sufficiently improve hearing. Although surgeries to improve hearing and active middle ear implants are both highly relevant in this patient population and both may be used for auditory rehabilitation in the same patient. There exist very few studies that compare the lifetime costs of these treatment options. Such studies are necessary for policymakers and surgeons to inform themselves about the cost efficacy of different interventions. In this paper we intend to fill this research gap. Methodology The analysis compared the costs of three care alternatives from the perspective of health insurers. In the following model, there are specific durations between the interventions. These are derived from the medical data. Within these time periods, the treatment success, e.g. hearing improvement, is defined as comparable between the options. The analysis does not include the results of hearing measurements (e.g. 799 Expensive today butcheaper tomorrow: lifetime costs ofanactive middle ear implant compared… in dB) or health-related quality of life. This assumption is essential for a meaningful comparison of lifetime costs: The following three groups of patients were defined: Group 1: Patients who received a middle ear implant directly, without prior hearing-improvement surgery. Group 2: Patients who underwent one or more hearing-improvement surgeries before receiving a middle ear implant. Group 3: Patients who had more than one hearingimprovement surgeries but no middle ear implant. The overall lifetime costs will be investigated considering several input parameters. A Monte Carlo simulation will be used as the methodological tool. In addition to the average cost per intervention, statements can be made about the benefit of the intervention based on the uncertainty of the individual input values. In particular, the parameters “implantation age/remaining lifetime” and the influence of the number of interventions will be looked at. Treatment data The treatment data in this retrospective study originated from the hospital information system of a maximum care hospital. Only the data of patients that gave their consent during treatment was analysed. Groups 1 and 2 consisted of all patients implanted with a VSB at the Hannover Medical School, Hannover, Germany, between 1997 and 2021, selected from the internal database for middle ear implants. Other types of middle ear implants were not included in this study. This resulted in 389 patients. Patients were excluded if their last appointment was earlier than 2019, as well as for having multimorbidity’s or private insurance. Subjects were selected and grouped based on: AMEI without prior corrective hearing-improvement surgery (group 1: n = 100) or corrective hearing-improvement surgery before AMEI implantation (group 2: n = 51). The patients were then screened so that each patient had an individual incident list with specific dates of their inpatient and outpatient visits related to their AMEI. The inpatient stays included appointments such as implantation, revision, or re-implantation. For group 2, the previous hearing-improvement operations were added. Outpatient visits included regular annual check-ups and processor fittings, visits for complications and/or hearing problems. These specific inpatient and outpatient events were then registered in SAP to confirm that the appointment in the internal database had indeed taken place and to specify which OPS and DRG codes of these appointments were used for billing purposes. For group 3, an SAP query was generated at the central department of MHH IT Services Applications, Enterprise Documents and Content, MHH Information Technology for patients with the OPS codes 5–195.90, .91, .92, .93 and 5–197.2. This determination of relevant patients in group 3 is done according to OPS codes (German procedure classification “Operationsund Prozedurenschlüssel-OPS”). In this step, group 2 serves as the basis for the selection of procedures. Based on the data in group 2, it was determined which hearing-improving surgeries are to be considered common before a VSB implant. The five OPS codes mentioned resulted from this analysis. Patients who received at least one of the procedures mentioned and were surgically treated between 2018 and 2021 and had statutory health insurance were included. Exclusion criteria were bilateral care and multimorbidity, since in these cases the data of the individual visit in SAP (especially for outpatient visits) could no longer be comprehensibly assigned to one specific pathology. This resulted in 211 patients, out of which 100 were randomly selected. Like the patients in groups 1 and 2, their visits were also registered in SAP so that an incident list with corresponding ICD codes, OPS codes and DRG codes could be created for each patient. Cost data Relevant cost data of the individual services were taken from the German health care system. For inpatient treatments, the hearing-improvement surgery costs were taken from the observed set of per-case flat rates (aG-DRG), including additional nursing revenues from the per-case flat rate catalogue [11]. For non-rated aG-DRGs, revenue values were used that were individually negotiated between health insurers and hospitals [2, 12–19]. These values also include the costs of the implant. For outpatient stays, the university’s outpatient flat rate of the hospital the data originated from was used. Consequently, all cost data are real data and not assumptions or estimates. Simulation The exported medical data was then validated, and the group classification was checked, as well as the age distribution between the groups. From the medical data, distribution identification was used to determine the theoretic distributions of the input parameters for individual model input parameters [20]. Normal distribution, truncated normal distribution, log-normal, exponential and uniform distribution were all tested. The patient’s age at first intervention, duration between hearing-improvement surgeries, duration between processor upgrades, and duration of care with hearing-improvement surgeries before VSB implantation were all considered. Furthermore, based on the observed probabilities for revision or re-implantation (depending on the duration of use) in VSB patients, a triangular distribution was calculated. 800 M.Krohn et al. We then performed a Monte Carlo simulation with the findings of the data preparation. The Monte Carlo simulation includes treatment data and cost data. Based on the individual input values and the identified distributions, the costs for all three patient groups were determined. For each group, 10,000 simulation runs were performed. The result of the simulation is thus not only a fixed individual result, but rather a probability distribution of the target values [21]. The aim is to ensure that the simulation model represents reality in the best possible way. The quality of the model is checked by comparing the model results with the treatment data from the hospital of origin. The individual steps of the simulation are explained below. The calculation is described in text form. The description covers the key aspects of the calculation. The factors described in text form are highlighted in bold in Tables1, 2, 3. The formal implementation is shown in Tables1, 2, 3. In addition to the aspects described, these tables contain all the calculation steps, including all the necessary intermediate steps and auxiliary calculations. Table4 shows the input parameters. Group 1: Patients who were directly fitted with a middle ear implant without first undergoing a hearing-improvement surgical intervention. Based on the distribution of age at initial intervention (“AFI”) and the life expectancy (“LEX”), which is defined as fixed, the treatment period (“TOP” = total observation period) was calculated. For this period, the costs of the first intervention (in this case, implantation of VSB) (“G1_CFI (group 1 - cost first intervention)”), the costs for re-implantations (“G1_CRI (group 1 - cost re-implantation)”), the costs for revisions (“G1_CRE (group 1 - cost revision)”), the costs for audio processor upgrades (“G1_APU (group 1 - audio processor upgrade)”) and the costs for an annual check-up were included for each simulation run. These add up to the lifetime cost of group 1 (“G1_LTC”). The costs for the first intervention include the costs of implantation (“ISB (Implantation Vibrant Soundbridge)”) based on the determined distribution and three outpatient follow-up appointments (“FOA (flat-rate outpatient appointments)”). The costs for re-implantations were similarly composed. The costs for “ISB” and “FOA” were also used here. The only difference is that the number of “FOA” appointments was reduced to two, since the third appointment was covered by the annual check-up appointment in the model. Furthermore, the time of a first or second re-implantation results from the lifetimes of the implants (“LTS (Lifetime Vibrant Soundbridge) or LTS1, LTS2 and LTS3”). The model was limited to a maximum of two re-implantations. Possible resulting model errors due to a possible increased number of required re-implantations was indicated by (“EP_LTS_G1 (error probability lifetime Vibrant Soundbridge group 1)”). This value thus partially describes the model quality. Furthermore, a discount rate (“DIR”) of 5% was applied for all costs arising in the future. This corresponds to the Hanoverian consensus and is taken into account in all parts of the model [22]. The number of implantations results from the initial implantation plus the number of necessary re-implantations and is defined as "NOI_G1 (number of implantations group 1)”. The costs for revisions include the costs of an inpatient revision stay (“REV”) according to the determined distribution plus an additional “FOA” appointment. The interval between necessary revisions is derived from the determined distribution. These distributions are defined as “TTR (time to revision) or TTR1, TTR2, TTR3, TTR4”. The model allows for a maximum of three revisions. The model error, which describes the need for additional revisions, is defined as “EP_TTR_G1 (error probability time to revision group 1)”. The number of revisions is defined as “NOR_G1 (number of revisions group 1)”. The cost of the audio processor upgrades was calculated from the cost of one audio processor (“APU”), taking into account the period between updates (“TAP”) and the total duration of care (“TOP”). The minimum time to upgrade is six years, as billing to health insurance is mainly possible after six years [23]. The time between the completed sixth year of use and the change to a new processor was derived from the observed data. This follows an exponential distribution.The number of processor upgrades is defined as “NOP_G1” (number of processor upgrades). The total cost of annual inspections (FOA) was calculated as the present value over the “TOP” period. Furthermore, the outputs “DR_LTS” and “DR_TTR” were generated. These describe the default rates of the implants (LTS) after 1, 10 and 25years of use and the revision probability (TTR) after 1, 5 and 10years. The values were used to determine the model quality. Here, the data from the hospital information system are compared with the results of the simulation runs. The outputs including all correction steps are shown in Table1. The outputs explicitly presented in the method description are printed in bold. Group 2: Patients who received a middle ear implant after one or more hearing-improvement middle ear surgeries. Group 2 patients underwent hearing-improvement surgeries prior to VSB implantation. Consequently, the first intervention consisted of a hearing-improvement surgery (“G2_ CFI (group 2-cost first intervention)”). Possible further hearing-improvement surgeries (“G2_CFS (group 2-cost further surgery) as well as the subsequent treatment with VSB have to be taken into account. This includes the costs of the first implantation (“G2_VSB (group 2-cost implantation VSB)”), the costs for re-implantations (“G2_CRI (group 2-cost re-implantation)”), the costs for revisions (“G2_CRE (group 2-cost revision)”) and the costs for audio processor 801 Expensive today butcheaper tomorrow: lifetime costs ofanactive middle ear implant compared… Table 1 Simulation model – group 1 Output Equation TOPuB1 if(round(LEX-AFIa) > (LEX-MIA),round(LEX-AFIb),round(LEX-AFIa)) TOPuB2 if(TOPuB1 > 77;round(LEX-AFIc);TOPuB1) TOPlB1 if(TOPuB2 < 1;round(LEX-AFIb);TOPuB2) TOPlB2 if(TOPlB1 < 1;round(LEX-AFIc);TOPlB1) TOPlB3 if(TOPlB2 < 1;1;TOPlB2) TOPuB3 if(TOPlB3 > (LEX-MIA);(LEX-MIA);TOPlB3) TOP TOPuB3 AFI LEX-TOP LTS1lB if(LTS1 < 0;0;LTS1) LTS2lB if(LTS2 < 0;0;LTS2) LTS3lB if(LTS3 < 0;0;LTS3) EP_LTS_G1 TOP-LTS1lB-LTS2lB-LTS3lB [upper spec limit = 0] DR_LTS1 round(LTS1lB-0,5) [lower spec limit = 1] DR_LTS10 round(LTS1lB-0,5) [lower spec limit = 10] DR_LTS25 round(LTS1lB-0,5) lower spec limit = 25] NOI_G1 1 + if(LTS1lB > = TOP;0;1) + if((LTS1lB + LTS2lB) > = TOP;0;1) TTR1lB if(TTR1 < 0;0;TTR1) TTR2lB if(TTR2 < 0;0;TTR2) TTR3lB if(TTR3 < 0;0;TTR3) TTR4lB if(TTR4 < 0;0;TTR4) TTR1uB TTR1lB TTR2uB TTR2lB TTR3uB TTR3lB TTR4uB TTR4lB EP_TTR_G1 TOP-TTR1uB-TTR2uB-TTR3uB-TTR4uB [upper spec limit = 0] DR_TTR1 round(TTR1uB-0.5) [lower spec limit = 1] DR_TTR5 round(TTR1uB-0.5) [lower spec limit = 5] DR_TTR10 round(TTR1uB-0.5) [lower spec limit = 10] NOR_G1 if(TTR1uB > = TOP;0;1) + if((TTR1uB + TTR2uB) > = TOP;0;1) + if((TTR1uB + TTR2uB + TTR3uB) > = TOP;0;1) TTU TTUa + TTUb TAP round(TTU) NOP_G1 if(TOP > = TAP;1;0) + if(TOP > = (2*TAP);1;0) + if(TOP > = (3*TAP);1;0) + if(TOP > = (4*TAP);1;0) + if(TOP > = (5*TAP);1 ;0) + if(TOP > = (6*TAP);1;0) + if(TOP > = (7*TAP);1;0) + if(TOP > = (8*TAP);1;0) + if(TOP > = (9*TAP);1;0) + if(TOP > = (10*TAP);1;0) + if(TOP > = (11*TAP);1;0) + if(TOP > = (12*TAP);1;0) G1_CFI (ISB + 3*FOA)/((1 + DIR)^0) G1_CRI if(LTS1lB < TOP;1;0)*((ISB + 2*FOA)/((1 + DIR)^(round(LTS1lB-0.5)))) + if((LTS1lB + LTS2lB) < TOP;1;0)*((ISB + 2*FOA)/ ((1 + DIR)^((round(LTS1lB + LTS2lB-0.5))))) G1_CRE if(TTR1uB < TOP;1;0)*((REV + FOA)/((1 + DIR)^(round(TTR1uB-0.5)))) + if((TTR1uB + TTR2uB) < TOP;1;0)*((REV + FO A)/((1 + DIR)^((round(TTR1uB + TTR2uB-0.5))))) + if((TTR1uB + TTR2uB + TTR3uB) < TOP;1;0)*((REV + FOA)/((1 + DIR )^((round(TTR1uB + TTR2uB + TTR3uB-0.5))))) G1_APU if(TOP > = TAP;1;0)*(APU/((1 + DIR)^TAP)) + if(TOP > = (2*TAP);1;0)*(APU/((1 + DIR)^(2*TAP))) + if(TOP > = (3*TAP); 1;0)*(APU/((1 + DIR)^(3*TAP))) + if(TOP > = (4*TAP);1;0)*(APU/((1 + DIR)^(4*TAP))) + if(TOP > = (5*TAP);1;0)*(APU/ ((1 + DIR)^(5*TAP))) + if(TOP > = (6*TAP);1;0)*(APU/((1 + DIR)^(6*TAP))) + if(TOP > = (7*TAP);1;0)*(APU/((1 + DIR)^( 7*TAP))) + if(TOP > = (8*TAP);1;0)*(APU/((1 + DIR)^(8*TAP))) + if(TOP > = (9*TAP);1;0)*(APU/((1 + DIR)^(9*TAP))) + if(TOP > = (10*TAP);1;0)*(APU/((1 + DIR)^(10*TAP))) + if(TOP > = (11*TAP);1;0)*(APU/((1 + DIR)^(11*TAP))) + if(TOP > = (12*TAP);1;0)*(APU/((1 + DIR)^(12*TAP))) G1_TSC FOA*((((1 + DIR)^(TOP-1))-1)/(((1 + DIR)^(TOP-1))*DIR)) G1_LTC G1_CFI + G1_CRI + G1_CRE + G1_APU + G1_TSC 802 M.Krohn et al. upgrades (“G2_APU (group 2 – audio processor upgrade)”). In addition, there are costs for an annual service appointment (“G2_TSC”) for the entire time. Together this results in the lifetime costs of group 2 (“G2_LTC”). During the period of treatment with hearing-improving surgeries, costs for possible hearing aids are not included. This point is addressed in the discussion. The content of G2_LTC is identical to G1_LTC. Furthermore, the contents of G2_VSB are also identical to G1_CRI, G2_CRI to G1_CRI, G2_APU to G1_APU and G2_CRE to G1_CRE. There are only two changes to be considered. The time of use of the VSB needs to be shortened because hearing-improvement operations were performed before the VSB. Furthermore, the number of FOA appointments is reduced to two for G2_VSB because the third appointment is included by G2_LTC. The previous observation time (“TOP”) is now split into “THI (time of hearing-improving interventions)” and “TSB (time of Vibrant Soundbridge)”. THI results from the determined distribution with a maximum value of 25years. This maximum definition is necessary because group 2 patients change to a VSB. If no maximum value were defined, treatment sequences could arise that correspond to group 3. The number of operations results from the determined success duration between hearingimprovement operations and THI. The duration of success of surgery was defined as “DSH1, DSH2, … (duration success hearing-improving surgery)”. The number of hearingimprovement surgeries was limited to six including initial intervention and is defined as NHI_G2. The probability of needing more surgeries is described as EP_NHI_G2 (error probability number of hearing-improvement surgeries group 2). The cost of hearing-improvement surgery was taken from the distribution of the observed aG-DRG set. Furthermore, two FOA appointments were included after each surgery. Table 2 Simulation model (additions) – group 2 Output Equation THIlB if(round(THIt) < TOP,round(THIt),TOP) THI if(THIlB < 25,THIlB,25) TSB TOP-THI NHI_G2 1 + if(DSH1 < THI,1,0) + if((DSH1 + DSH2) < THI,1,0) + if((DSH1 + DSH2 + DSH3) < THI,1,0) + if((DSH1 + DSH2 + DSH3 + DSH4) < THI,1,0) + if((DSH1 + DSH2 + DSH3 + DSH4 + DSH5) < THI,1,0) EP_NHI_G2 THI-DSH1-DSH2-DSH3-DSH4-DSH5-DSH6 G2_CFI (HIS1 + 2*FOA)/((1 + DIR)^0) G2_CFS if(DSH1 < THI,1,0)*((HIS2 + FOA)/((1 + DIR)^round(DSH1-0,5))) + if((round(DSH1 + DSH2)) < THI,1,0)*((HIS3 + FOA)/((1 + DIR)^round(DSH1 + DSH2-0,5))) + if((round(DSH1 + DSH2 + DSH3)) < THI,1,0)*((HIS4 + FOA)/ ((1 + DIR)^round(DSH1 + DSH2 + DSH3-0,5))) + if((round(DSH1 + DSH2 + DSH3 + DSH4)) < THI,1,0)*((HIS5 + FOA)/ ((1 + DIR)^round(DSH1 + DSH2 + DSH3 + DSH40,5))) + if((round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5)) < THI,1,0)*((HIS6 + FOA)/ ((1 + DIR)^round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5-0,5))) G2_VSB (ISB + 2*FOA)/((1 + DIR)^THI) G2_CRI if(LTS1lB < TSB;1;0)*((ISB + 2*FOA)/((1 + DIR)^round(THI + LTS1lB-0,5))) + if((LTS1lB + LTS2lB) < TSB;1;0)*((ISB + 2*F OA)/((1 + DIR)^round(THI + LTS1lB + LTS2lB-0,5))) NOI_G2 1 + if(LTS1lB > = TSB,0,1) + if((LTS1lB + LTS2lB) > = TSB,0,1) G2_CRE if(TTR1uB < TSB,1,0)*((REV + FOA)/((1 + DIR)^round(THI + TTR1uB-0,5))) + if((TTR1uB + TTR2uB) < TSB,1,0)*((REV + F OA)/((1 + DIR)^round(THI + TTR1uB + TTR2uB-0,5))) + if((TTR1uB + TTR2uB + TTR3uB) < TSB,1,0)*((REV + FOA)/((1 + DIR)^round(THI + TTR1uB + TTR2uB + TTR3uB-0,5))) NOR_G2 if(TTR1uB > = TSB,0,1) + if((TTR1uB + TTR2uB) > = TSB,0,1) + if((TTR1uB + TTR2uB + TTR3uB) > = TSB,0,1) G2_TSC G1_TSC G2_APU if(TSB > = TAP;1;0)*(APU/((1 + DIR)^(TAP + THI))) + if(TSB > = (2*TAP);1;0)*(APU/((1 + DIR)^(2*TAP + THI))) + if(TSB > = (3*TAP);1;0)*(APU/((1 + DIR)^(3*TAP + THI))) + if(TSB > = (4*TAP);1;0)*(APU/((1 + DIR)^(4*TAP + THI))) + if(TSB > = (5*TAP);1;0)*(APU/((1 + DIR)^(5*TAP + THI))) + if(TSB > = (6*TAP);1;0)*(APU/((1 + DIR)^(6*TAP + THI))) + if(TSB > = (7*TAP);1;0)*(APU/((1 + DIR)^(7*TAP + THI))) + if(TSB > = (8*TAP);1;0)*(APU/((1 + DIR)^(8*TAP + THI))) + if(TSB > = (9*TAP);1;0)*(APU/((1 + DIR)^(9*TAP + THI))) + if(TSB > = (10*TAP);1;0)*(APU/((1 + DIR)^(10*TAP + THI))) + if(T SB > = (11*TAP);1;0)*(APU/((1 + DIR)^(11*TAP + THI))) + if(TSB > = (12*TAP);1;0)*(APU/((1 + DIR)^(12*TAP + THI))) NOP_G2 if(TSB > = TAP;1;0) + if(TSB > = (2*TAP);1;0) + if(TSB > = (3*TAP);1;0) + if(TSB > = (4*TAP);1;0) + if(TSB > = (5*TAP);1; 0) + if(TSB > = (6*TAP);1;0) + if(TSB > = (7*TAP);1;0) + if(TSB > = (8*TAP);1;0) + if(TSB > = (9*TAP);1;0) + if(TSB > = ( 10*TAP);1;0) + if(TSB > = (11*TAP);1;0) + if(TSB > = (12*TAP);1;0) G2_LTC G2_CFI + G2_CFS + G2_VSB + G2_CRI + G2_CRE + G2_APU + G2_TSC EP_LTS_G2 TSB-LTS1lB-LTS2lB-LTS3lB EP_TTR_G2 TSB-TTR1uB-TTR2uB-TTR3uB-TTR4uB 803 Expensive today butcheaper tomorrow: lifetime costs ofanactive middle ear implant compared… The outputs and equations including all correction steps are shown in Table2. The outputs explicitly presented in the method description are printed in bold. Group 3: Patients who received more than one hearingimprovement surgery but no middle ear implant. Patients in group 3 received hearing-improvement surgery throughout the observation period. The costs of this group consist of the costs of the first intervention (“G3_ CFI”) plus the costs of further hearing-improvement surgeries (“G3_CFS”), as well as an assumed annual control visit whose total costs are represented by “G3_TSC”. G3_TSC corresponds to G1_TSC in terms of content. Together this results in the lifetime costs of group 3 (“G3_LTC”). The time between hearing-improvement operations corresponds to the value “DSH1, DSH2, …” already presented in group 2. The number of interventions, whose costs per intervention follow the same distribution as in group 2 (“HIS”), is limited in the model to 15 interventions (initial intervention plus a maximum of 14 follow-up interventions). The probability of higher need is described by EP_NHI_G3 (error probability number of hearing-improvement surgeries group 3). Consequently, group 3 corresponds to group 2 in terms of content, but the provision of hearing-improvement surgery takes place over the entire duration (“TOP”), not only during a short period (“THI”) before a subsequent VSB implantation. The outputs including all correction steps are shown in Table3. The outputs explicitly presented in the method description are printed in bold. In addition, it should be noted for group 3 that no costs are included for the provision of hearing aids. This point is addressed in the discussion. Model parameters The following Table4 shows the input parameters of the simulation model. The source of the data origin is described in the last column. Table 3 Simulation model (additions) – group 3 Output Equation G3_CFI (HIS1 + 2*FOA)/((1 + DIR)^0) G3_CFS if(DSH1 < TOP;1;0)*((HIS2 + FOA)/((1 + DIR)^round(DSH1-0.5))) + if((round(DSH1 + DSH2)) < TOP;1;0)*((HIS3 + FOA)/((1 + DIR)^round(DSH1 + DSH2-0.5))) + if((round(DSH1 + DSH2 + DSH3)) < TOP;1;0)*((HIS4 + FOA)/ ((1 + DIR)^round(DSH1 + DSH2 + DSH3-0.5))) + if((round(DSH1 + DSH2 + DSH3 + DSH4)) < TOP;1;0)*((HIS5 + FOA)/ ((1 + DIR)^round(DSH1 + DSH2 + DSH3 + DSH4-0.5))) + if((round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5)) < TOP;1;0)*((HIS6 + FOA)/ ((1 + DIR)^round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5-0.5))) + if((round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6)) < TOP;1;0)*((HIS7 + FOA)/ ((1 + DIR)^round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH60.5))) + if((round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7)) < TOP;1;0)*((HIS8 + FOA)/((1 + DIR)^round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH70.5))) + if((round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8)) < TOP;1;0)*((HIS9 + FOA)/ ((1 + DIR)^round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH80.5))) + if((round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9)) < TOP;1;0)*(( HIS10 + FOA)/((1 + DIR)^round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH90.5))) + if((round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH10)) < TOP;1;0)*(( HIS11 + FOA)/((1 + DIR)^round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH100.5))) + if((round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH10 + DSH11)) < TOP;1;0)*(( HIS12 + FOA)/((1 + DIR)^round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH10 + DSH110.5))) + if((round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH10 + DSH11 + DSH12)) < TOP;1;0)*(( HIS13 + FOA)/((1 + DIR)^round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH10 + DSH11 + DSH120.5))) + if((round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH10 + DSH11 + DSH12 + DSH13)) < TOP;1;0)*(( HIS14 + FOA)/((1 + DIR)^round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH10 + DSH11 + DSH12 + DSH130.5))) + if((round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH10 + DSH11 + DSH12 + DSH13 + DSH14)) < TOP;1;0)*((HIS15 + FOA)/ ((1 + DIR)^round(DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH10 + DSH11 + DSH12 + DSH13 + DSH14-0.5))) G3_TSC G1_TSC EP_NHI_G3 TOP-DSH1-DSH2-DSH3-DSH4-DSH5-DSH6-DSH7-DSH8-DSH9-DSH10-DSH11-DSH12-DSH13-DSH14-DSH15 G3_LTC G3_CFI + G3_CFS + G3_TSC NHI_G3 1 + if(DSH1 < TOP;1;0) + if((DSH1 + DSH2) < TOP;1;0) + if((DSH1 + DSH2 + DSH3) < TOP;1;0) + if((DSH1 + DSH2 + DSH3 + DSH4) < TOP;1;0) + if((DSH 1 + DSH2 + DSH3 + DSH4 + DSH5) < TOP;1;0) + if((DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6) < TOP;1;0) + if((DSH1 + DSH2 + DSH3 + DSH4 + D SH5 + DSH6 + DSH7) < TOP;1;0) + if((DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8) < TOP;1;0) + if((DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9) < TOP;1;0) + if((DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH10) < TOP;1; 0) + if((DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH10 + DSH11) < TOP;1;0) + if((DSH1 + DSH2 + DSH3 + DSH4 + D SH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH10 + DSH11 + DSH12) < TOP;1;0) + if((DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + D SH9 + DSH10 + DSH11 + DSH12 + DSH13) < TOP;1;0) + if((DSH1 + DSH2 + DSH3 + DSH4 + DSH5 + DSH6 + DSH7 + DSH8 + DSH9 + DSH10 + DSH11 + DSH12 + DSH13 + DSH14) < TOP;1;0) 804 M.Krohn et al. Results Treatment duration (“TOP”) In the model, the duration of care was calculated from the difference between a fixed life expectancy of 81years and the age at first intervention according to distribution identification. This resulted in a “TOP” duration of 26.73years, with a standard deviation of 13.94years. This duration describes the period of lifetime costs. In group 1 it includes the care with VSB, and in group 3 the care by means of hearingimprovement operations. For group 2, this period was divided into the care by means of hearing-improvement operations (“THI”) and the treatment time with VSB (“TSB”). Figure1 shows the distribution for “TOP”. In group 2, the duration of care with hearing-improvement surgery was 4.01years (± 5.18), and for care with VSB 22.72years (± 14.60). The sum of the durations again corresponds to TOP. Time in years is plotted on the x-axis, and the number of observations in the model is plotted on the y-axis. Model results ofgroup 1 In group 1, care was provided exclusively with a VSB. During the 26.73 [± 13.94]-year treatment period, a mean of 1.15 [± 0.39] implantations were performed. The number of re-implantations over the entire lifetime is therefore 0.15, with 86.6% of patients requiring no re-implantation and 12.2% requiring only one re-implantation. The number of revision surgeries necessary over the entire period of care averaged 0.45 (± 0.70), with 65.6% of patients requiring no revision and 26.0% requiring only one revision. The average number of processor upgrades was 3.64 (± 2.19). The values shown relate to the overall model. Due to the distribution of “TOP”, 50% of patients in the model will have a VSB treatment duration of more than 25years. The following figures show the number of implantations (Fig.2) and revisions (Fig.3). The question arises as to whether the model can represent reality. To answer this, the comparison between model results and real data must be used. The model shows the following values for the failure probabilities of the implants (DR_LTS). Table 4 Input values Input name Distribution Parameters Source LEX – life expectancy Fixed Value: 81 [24] (mean value man and woman) MIA – minimum implantation age Fixed Value: 5 [25] AFIa, AFIb, AFIc – age at first intervention (for lower/ upper bound correction) Weibull Shape: 244.32323 Scale: 2866.73542 Threshold: -2806.08613 own, based on medical data FOA – flat-rate outpatient appointments Fixed Value: 145 own, based on MHH cost data DIR – discount rate Fixed Value: 0.05 [22] ISB – cost of implantation of Vibrant Soundbridge Lognormal Location: 9.57283 Scale: 0.07429 own, based on medical data and cost data mentioned before LTS1, LTS2, LTS3 – lifetime Vibrant Soundbridge Triangular Lower: -3 Mode: 0 Upper: 400 own, based on medical data TTR1, TTR2, TTR3 – time to revision Triangular Lower: -3 Mode: 0 Upper: 140 own, based on medical data REV – cost revision Lognormal Location: 8.29047 Scale: 0.20121 own, based on medical data and cost data set APU – cost audio processor Fixed Value: 5719.9 [26] TTUa – time to upgrade APU minimum Fixed Value: 6 [23] TTUb – additional time to upgrade APU Exponential Rate: 1.41988 own, based on medical data THIt – Time of hearing-improving intervention total (for lower/upper bound correction) Lognormal Location: 0.70939 Scale: 1.3453 own, based on medical data DSH1, DSH2, … DSH15 – duration of success of hearing-improving surgery Weibull Shape: 0.68361 Scale: 1.74601 Threshold: 0.28273 own, based on medical data HIS1, HIS2, … HIS15 – cost of hearing-improving surgery Lognormal Location: 8.2339 Scale: 0.25662 own, based on medical data and cost data set 811 Expensive today butcheaper tomorrow: lifetime costs ofanactive middle ear implant compared… cannot be treated for the remaining time. Thus, from a cost perspective, alternative treatment options must be considered. The VSB seems to be a reasonable choice due to its lower lifetime costs. Homogeneity ofcosts With respect to cost homogeneity, the results from the entire simulation as well as the costs for remaining lifetimes of 10, 20 and 30 years will be examined as examples. Table5 shows the mean costs as well as the homogeneities of the costs. A t-test was used to check whether the mean values are different and if significant differences existed in the standard deviation. A p-value of less than 0.05 indicates significant difference. The costs of group 1 and 3 did not differ throughout the entire simulation. This is due to the fact that for short treatment durations G1 dominates, but for long durations group 3 is preferable, as was previously stated. Nevertheless, the differences in the standard deviations were always significant-and thus significantly higher in groups 2 and 3 due to the influence of the hearing-improvement operations. Consequently, treatment with the VSB had significantly higher homogeneity coefficients, attributable to very low midand long-term costs, due to small numbers of revision surgeries and re-implantations over the entire period of care. Therefore, the use of the middle ear implant therefore reduces the risk of cost outliers, even though the implantation of an active middle ear implant is often associated with increased complexity and higher costs [27]. The advantage of the lower standard deviation is evident in Fig.11. Costs in group 3 scatter significantly more than in the other groups. Furthermore, it should be noted that the analysis presented here excludes the costs of hearing aid fitting, which is often necessary after hearing-improvement surgeries. Hence, lifetime costs of hearing aid fitting, which can sum up to 4518€, as determined by Thum etal. should have been taken into account either in order to comprehensively report total lifetime costs of group 3 [28]. Also, additional treatment costs for the more than 25% of patients in group 3 who couldn’t be successfully treated (e.g. due to ear infections) until the end of life, would have been needed to be considered, too. These 25% of patients could potentially be described as a subset of patients with insufficient benefit from this combination therapy, as even state-of the-art power hearing aids would not be able to provide sufficient gain and/or output required by these patients [29]. For these 25% of patients who couldn’t be successfully treated in group 3, bypassing the middle ear and air–bone gap (ABG) with a powerful active middle ear implant or a direct acoustic stimulation of the cochlea, might have been the more promising hearing solution [29]. From an economic point of view, the following aspect can be added: Fig. 10 Distribution of successfully treated years in the simulation model. (source: own simulation) 812 M.Krohn et al. – The use of a middle ear implant makes future costs significantly easier to plan, since there is a higher degree of cost homogeneity. Influence ofthenumber ofhearing‑improvement surgeries Another question to be addressed regards the influence of the number of hearing-improvement surgeries administered. We aimed to find out whether the average costs depend on the number of hearing-improvement operations. Table6 shows the mean costs and standard deviations. A t-test was used to determine significance. A p value of less than 0.05 indicates that the costs were significantly increased compared to “one surgery less”. The costs increase significantly as the number increases. In the simulation, the difference between two and three hearing-improvement operations is the smallest. This is due to the age distribution in the simulation. Here it becomes clear that with a relatively low remaining life expectancy, an additional third surgery can postpone VSB implantation so long that no processor upgrade will be necessary shortly before death. However, it has already been shown in the previous discussion points that group 3 seems to be advantageous at low life expectancies. From an economic perspective, the following can be stated: switching to VSB implantation seems to be reasonable even after hearing-improvement surgeries have already been performed, because each subsequent hearingimprovement surgery significantly increases the lifetime costs. Figure12 shows the distribution functions, which is helpful for discussion of more than just the mean values. These distribution functions were determined by distribution identification from the simulation results. “0” – G1_LTC – Gen. Logistick=0.06296s=3351.4 m=27,976.0 “1” – Gen. Logistick=0.04093s=3582.5m=30,364.0 “2” – Burrk=2.0806a=7.2139b=36,652.0 “3” – Burr k=3.1797a=6.2471b=41,144.0 “4” – Weibull (3P)a=2.9827b=23,244.0g=13,307.0 “5” – Johnson SBg=2.3776d=3.3915l=1.1399E+5 x=−2760.1 In about 80% of cases, fewer hearing-improvement operations incur less costs. Patients with five hearing-improvement operations always result in the highest costs. In approximately 25% of cases, the difference between two, three or four hearing-improvement operations is unclear. This effect again mainly results in patients with a low remaining life expectancy, where a change to the VSB is not cost-effective. However, one hearing-improvement operation always appears to be better than two or more. This option is dominated only by the direct VSB option. From an economic point of view, it can be concluded that if the remaining life expectancy is sufficient, a switch to VSB treatment after the first hearing-improvement surgery seems to be reasonable. Based on the previous aspects of the discussion, it can be concluded: 1) When the life expectancy is less than 14years, the most favorable is a hearing-improving surgery. 2) In patients with residual lifetimes of more than 35years, the success of exclusive treatment with hearingimprovement surgery is unlikely. Since it is shown that a low number of hearing-improvement operations results in lower lifetime costs, direct VSB implantation appears to be a good choice. 3) If the remaining lifespan is between 14 and 36years, a definitive statement about the advantageousness is difficult, but results show that downscaling hearingimproving operations makes sense. The following is an example of a possible economic recommendation. Table 5 Costs and cost homogeneities depending on the duration of treatment. (source: own simulation) Total simulation set (TOP) 10years 20years 30years n in simulation 10,000 196 323 258 Group 1 Average cost in € 28,325 21,123 27,037 30,842 Standard deviation in € 6121 3249 4258 4652 Cost homogeneity 0.82 0.87 0.86 0.87 Group 2 Average cost in € 32,187 25,321 31,096 34,232 Standard deviation in € 6902 4496 5662 5975 Cost homogeneity 0.82 0.85 0.85 0.85 Group 3 Average cost in € 28,381 18,471 27,409 32,552 Standard deviation in € 10327 8018 9690 8037 Cost homogeneity 0.73 0.70 0.74 0.80 t-test–mean cost G1 vs. G2 < 0.001 < 0.001 < 0.001 < 0.001 G1 vs. G3 0.637 < 0.001 0.528 0.003 G2 vs. G3 < 0.001 < 0.001 < 0.001 0.007 Bonnet-test-standard deviation G1 vs. G2 < 0.001 0.013 0.002 0.031 G1 vs. G3 < 0.001 < 0.001 < 0.001 < 0.001 G2 vs. G3 < 0.001 < 0.001 < 0.001 < 0.001 813 Expensive today butcheaper tomorrow: lifetime costs ofanactive middle ear implant compared… 1. If TOP < 14 (age at first intervention > 67), then G3 (only hearing-improvement surgery) 2. If TOP > 35 (age at first intervention < 46), then direct G1 (VSB implantation) 3. If 13 < TOP < 36, then hearing-improvement surgery (age at first intervention 46 to 67). Change to VSB after one surgery if remaining lifetime at date of new intervention > 13years (corresponds to maximum age of 67) (= G2 with hearing-improving surgery), otherwise remain at G3 (only hearing-improvement surgery). Based on the age distribution of TOP, we have four different patient groups. 16.46% of patients can be exclusively treated with hearing-improvement surgeries due to their age at first intervention (> 67). 5.93% of patients also receive only hearing-improvement surgeries due to their age at second invention being 68 or older. 23.76% of patients can undergo direct VSB implantation because their age at first intervention is less than 46years. 53.85% of patients can first receive hearing-improvement surgery, followed by a VSB. Such a strategy leads to a mean cost of 29,099.50€ with a standard deviation of 8384.29€. This is still significantly above the cost of direct VSB implantation, but below the cost of hearing-improvement surgery followed by VSB implantation (each p < 0.001). The main effect results in the number of years the patient is successfully treated. The costs per successfully treated year of life are 1088.61€. This is still higher than the cost of VSB (1059.61€) per year, but the model also only considers cases in Fig. 11 Overview of simulation results (here: G1 green; G2 red; G3 blue) (source: own simulation) Table 6 Influence of the number of hearing-improvement operations on lifetime costs (source: own simulation) NHI (G1 or G2 data) Mean in € Standard deviation Cost increased? “0” = G1_LTC 28,325 6121 – “1” 30,606 6499 < 0.001 “2” 32,347 6493 < 0.001 “3” 32,796 6953 0.030 “4” 34,067 7567 < 0.001 “5” 35,394 7361 < 0.001 814 M.Krohn et al. which a hearing-improvement surgery is followed by a second intervention. However, the costs are reduced by more than 10% compared to the other strategies. Conclusion Choosing a treatment option is a complex medical decision. Treatment success and outcome for the patient is difficult to assess from the “economists’ desk”. Cost data, probabilities from medical data sets and knowledge of treatment processes enable us to estimate future costs. This can support medical decision-making, which is also often influenced by short-term cost aspects. Our study has shown that time plays a decisive role. We found that differentiated decision making can reduce lifetime costs, and treatment alternatives perceived as costly today may be beneficial in the long run. Direct, first-line VSB implantation or implantation after a low number of hearing-improvement surgeries is a reasonable treatment option. This opens up the possibility of a stronger patient-centered approach in decision-making. When focusing on the current costs of a single intervention, often only the option of a (further) hearing-improving surgery remained open due to the refusal of coverage by health insurance. However, if the total lifetime costs are considered, a broader set of treatment alternatives (direct implantation or first hearing-improving surgery) may result for the patient (taking into account his or her age), which can also be considered economically advantageous. From an economic point of view – in particular from the point of view of health insurance companies – our analysis shows that the determination of lifetime costs is an important basis for economic decision-making that goes beyond the individual financial year of the health insurance company. Based on patient data, it was clearly shown that it is economically feasible to consider the option of VSB implantation earlier. With regard to the methodology of our analysis, it should be noted that cost analyses are always subject to the limitation that the input parameters can change or that different health care systems may have different cost structures. Furthermore, data on events in the distant future, whether medical or economic, is often subject to uncertainties. Future events are therefore difficult to calculate. However, the uncertainty can be reduced by using simulation models that take distribution functions into account. It should also be noted that a conservative approach in the inclusion of unclear events such as complications reduces uncertainty. The focus of this analysis was on the cost of treatment. Although “successfully treated years” were addressed based on the times between surgeries, the quality of care in these years was not considered. Other methods of health economic Fig. 12 Cumulative distribution functions of the costs for the treatment alternatives “0 to 5 hearing-improvement operations prior to VSB implantation” (source: own simulation) 815 Expensive today butcheaper tomorrow: lifetime costs ofanactive middle ear implant compared… evaluation relate treatment outcomes to costs. Conceivable outcomes include factors such as decibels (dB) or gains in quality of life (e.g., QALYs). What these analyses have in common is that knowledge about costs is necessary. Thus, the present results can be regarded as a starting point for further analyses, methodologically as well as for the current object of investigation, VSB implantation. Author contributions MK developed the methodology and wrote the first draft of the manuscript. His focus was on the methodology, the results and the discussion. KK had the project lead, wrote the background and supported the development of the methodology and discussion. KK was supported in this by BS and CB. AB coordinated the project, exported the medical data and assisted with the analysis and background. SB, TL, ALS and HM ensured the maintenance and expansion of the medical database and the creation of export options as well as the possibility of medical discussion approaches. SF, ALS, CB and MU had the research idea and led the project in the different dimensions of expertise required. All authors read and approved the final manuscript. Funding Open Access funding enabled and organized by Projekt DEAL. Funding information can be found in the ‘Conflict of interest’ section. Availability of data and materials The data that support the findings of this study are available from the corresponding author, upon reasonable request. Declarations Conflict of interest This study was partly financed by MED-EL. The authors confirm that the results of this analysis were not influenced at all by this sponsorship. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. References 1. RI Innovation GmbH: DRG D30B. https:// app. reimb ursem ent. info/ drgs/ D06B?s= D06B (2023) 2. Charité: Entgelttarif für Krankenhäuser im Anwendungsbereich des Krankenhausentgeltgesetzes (KHEntgG) und der Bundespflegesatzverordnung (BPflV) sowie Unterrichtung der Patientinnen und Patienten gemäß § 8 KHEntgG / § 8 BPflV. https:// www. chari te. de/ filea dmin/ user_ upload/ portal/ klini kum/ behan dlung_ stati onaer/ Entge lttar if. pdf (2023) 3. Chadha, S., Kamenov, K., Cieza, A.: The world report on hearing, 2021. Bull World Health Organ (2021). https:// doi. org/ 10. 2471/ BLT. 21. 285643 4. Neubauer, G., Gmeiner, A.: Volkswirtschaftliche Bedeutung von Hörschäden und Möglichkeiten zur Reduktion deren Folgekosten. https:// www. kbv. de/ media/ sp/ 120811_ studie_ ifg_ absch lussb ericht. pdf (2011) 5. Schield, B.: Hearing Loss – Numbers and Costs. EVALUATION OF THE SOCIAL AND ECONOMIC COSTS OF HEARING IMPAIRMENT A report for Hear-It AISBL. https:// www. ehima. com/ wpconte nt/ uploa ds/ 2021/ 01/ HearitRepor tHear i ngLossNumbe rsandCosts2019. pdf (2019) 6. Institute for Health Metrics and Evaluation (IHME): GBD Compare. http:// vizhub. healt hdata. org/ gbdcompa re (2015). Accessed 15 Jun 2024 7. Löhler, J., Walther, L.E., Hansen, F., Kapp, P.C., Meerpohl, J.J., Wollenberg, B., Schönweiler, R., Schmucker, C.: The prevalence of hearing loss and use of hearing aids among adults in Germany. A systematic review. Eur Arch Otorhinolaryngol Head Neck 276, 945–956 (2019) 8. Schmucker, C., Kapp, P.C., Motschall, E., Löhler, J., Meerpohl, J.J.: Prevalence of hearing loss and use of hearing aids among children and adolescents in Germany: a systematic review. Universität, Freiburg (2019) 9. von Gablenz, P., Holube, I.: Prävalenz von Schwerhörigkeit im Nordwesten Deutschlands: Ergebnisse einer epidemiologischen Untersuchung zum Hörstatus (HÖRSTAT). HNO (2015). https:// doi. org/ 10. 1007/ s001060142949-7 10. Dispenza, F., Battaglia, A.-M., Salvago, P., Martines, F.: Determinants of failure in the reconstruction of the tympanic membrane: a case-control study. Iran J Otorhinolaryngol 30, 341–346 (2018) 11. InEK: Institut für das Entgeltsystem im Krankenhaus - Fallpauschalen-Katalog 2023. https:// www.gdrg. de/ agdrgsystem2023/ fallp ausch alenkatal og/ fallp ausch alenkatal og20232 (2022) 12. Universitätsklinikum Erlangen: Fd-91–01 DRG-Entgelttarif und Pflegekostentarif 2023. erlangen.de/fileadmin/dateien/patienten/ patienten_a-z/Pflegekostentarif_DRG.pdf (2022) 13. Katholisches Klinikum Bochum: DRG-Entgelttarif im Anwendungsbereich des KHEntgG Unterrichtung des Pateinten gemäß §8 KHEntgG. https:// www. klini k umbochum. de/ avb. html? file= files/ klini kumbochum/ dokum ente/ AVB/ 202309% 20DRGEntge lttar if% 20inkl. MLK. pdf (2023) 14. Klinikum Nürnberg: Entgelttarif des Klinikum Nürnberg und Unterrichtung der Patienten gemäß § 8 Abs. 9 KHEntgG bzw. § 8 Abs. 5 BPflV. https:// www. klini kumnuern berg. de/ filea dmin/ user_ upload/ Fl yer_ Dokum ente_ Verw a ltung/ P atie ntena ufnah me_ und_ Leis t ungsa brech nung/ Entge lttar if_ 01. 11. 2023_ Versi on_1. 0_ feu_ 25102 023. pdf (2023) 15. Klinikum Stuttgart: DRG-Entgelttarif (gemäß Krankenhausentgeltgesetz (KHEntgG)) und PEPP-Entgelttarif (gemäß Bundespflegesatzverordnung (BPflV)). https:// www. klini kumstutt gart. de/ filea dmin/ media pool/ do wnl oads/ 230220_ Entge ltTar if_ Klini kumSt uttg a rt_m_ Anlag en. pdf (2023) 16. Medizinische Hochschule Hannover: DRG und PEPP Entgeltund Behandlungskostentarif Unterrichtung des Patienten gemäß § 8 Abs. 8 KHEntgG und § 8 Abs. 5 BPflV. https:// www. mhh. de/ filea dmin/ mhh/ gbklini kmana g eme nt/ downl oads/ 210105_ DR GEntg e lt-_ und_ Behan dlung skost entar if_ ohne_ Fussz eile. pdf (2021) 816 M.Krohn et al. 17. Universitätsklinikum Bonn: Entgelttarif 2021 Unterrichtung des Patienten gemäß § 8 KHEntgG. https:// www. ukbonn. de/ site/ assets/ files/ 18552/ drg_ entge lttar if_ 01072 021. pdf (2021) 18. Universitätsklinikum Ulm: aDRG Entgelttarif für das Universitätsklinikum Ulm Unterrichtung des Patienten gemäß § 8 KHEntgG Abs.8 (Krankenhausentgeltgesetz). http:// www. unive rsita etskl inikulm. de/ filea dmin/ def au lt/ 05_ Uberuns/ DR G_ Entge lttar if_ fuer_ das_ Unive rsita etskl inikum_ 01_ 2020_ Versi on_2. pdf (2020) 19. Universitätsmedizin Göttingen: Krankenhausentgelttarif gültig ab 01.01.2022 für stationäre und teilstationäre Behandlung in der Universitätsmedizin Göttingen (2022) 20. Harrell, C., Ghosh, B.K., Bowden, R.: Simulation using ProModel. McGraw-Hill series in industrial engineering and management science. McGraw-Hill, Boston, Mass. (2000) 21. Fleßa, S.: Systemisches Krankenhausmanagement, 2nd edn. De Gruyter Studium, Berlin (2022) 22. von der Graf, S.J.M., Greiner, W., Jost, F., Klusen, N., Kubin, M., Leidl, R., Mittendorf, T., Rebscher, H., Schöffski, O., Vauth, C., Volmer, T., Wahler, S., Wasem, J., Weber, C.: Deutsche Empfehlungen zur gesundheitsökonomischen Evaluation - dritte und aktualisierte Fassung des Hannoveraner Konsens. Gesundheitsökonomie Qualitätsmanagement (2007). https:// doi. org/ 10. 1055/s2007963505 23. GKV-Spitzenverband: Apparative Versorgung bei Schwerhörigkeit. Begutachtungsanleitung (2020) 24. Statistisches Bundesamt: Sterbefälle und Lebenserwartung. https:// www. desta tis. de/ DE/ Themen/ Gesel lscha ftUmwelt/ Bevoe lkeru ng/ Sterb efael leLeben serwa rtung/_ inhalt. html (2023) 25. MED-EL Medical Electronics: VIBRANT SOUNDBRIDGE Middle Ear Implant. https:// www. medel. com/ heari ngsolut ions/ vibra ntsound bridge (2023) 26. Donau Vienna Insurance Group: Versicherung ausschließlich für Hörsysteme der Firma MED-EL. https:// www. vbk. co. at/ wpconte nt/ uploa ds/ 2023/ 04/ MEDELAntrag042023. pdf (2022) 27. Wardenga, N., Diedrich, V., Waldmann, B., Lenarz, T., Maier, H.: Hearing aid treatment in patients with mixed hearing loss. Part I: expected benefit and limitations after stapes surgery. Audiol Neurotol (2020). https:// doi. org/ 10. 1159/ 00050 2994 28. Thum, C., Langenfeld, R., Flessa, S.: Lebenszeitkosten der Hörgeräteversorgung in Deutschland (lifetime cost of hearing aids in Germany). Laryngorhinootologie (2021). https:// doi. org/ 10. 1055/a12538325 29. Wardenga, N., Snik, A.F.M., Kludt, E., Waldmann, B., Lenarz, T., Maier, H.: Hearing aid treatment for patients with mixed hearing loss. Part II: speech recognition in comparison to direct acoustic cochlear stimulation. Audiol Neurotol (2020). https:// doi. org/ 10. 1159/ 00050 4285 Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.