Citation: Siegel, E., Luecke, T., Vogelmann, T.: “Costs and resource consumption of patients with painful diabetic polyneuropathy”, in: “Monitor Healthcare Research” (06/22), pp. 61-67. http://doi.org/10.24945/MVF.06.22.1866-0533.2464
Plain Text
Costs and resource utilization of patients with painful diabetic polyneuropathy
Diabetic polyneuropathy (DPN) is associated with significantly increased healthcare costs and thus an economic burden on society and the healthcare system (Happich et al. 2008; Khdour 2020; Shillo et al. 2019). A high number of hospital stays, need for care, incapacity to work as well as follow-up costs of DPN (e.g. amputations) and an overall greater use of healthcare resources are responsible for the higher medical costs (Sadosky et al. 2015; Schroeter/Hartung 2005; Ziegler 2020) and illustrate this Burden on patients suffering from DPN. The costs tend to increase as the burden of illness increases (Happich et al. 2008).
>> There are currently no current economic key figures for patients with DPN for Germany, in particular a distinction between s-DPN and non-s-DPN. This study aims to analyze the service expenditure including sickness benefit payments and days of incapacity for work of patients with DPN (painful/non-painful) from the perspective of statutory health insurance in Germany.
methodology
Study design and data source
A retrospective analysis of routine data from those insured by statutory health insurance (GKV) was carried out. Anonymized, age- and gender-adjusted secondary data were used for the analysis. The research database used contains anonymized data on around 4.9 million people insured by statutory health insurance from various statutory health insurance companies (Andersohn/Walker 2016) and represents around 7% of people with statutory health insurance (2019: 73.5 million (Federal Ministry of Health 2021)) in Germany. In addition to master data such as age, gender or insurance status, the database contains diagnoses - encrypted according to the International Statistical Classification of Diseases and Related Health Problems, 10th Revision, German Modification (ICD-10-GM) - from the outpatient and inpatient sectors. Information about drug prescriptions in the outpatient sector is available encrypted using codes from the anatomical therapeutic chemical classification (ATC) or the pharmaceutical central number (PZN). Treatments, such as operations or diagnostic procedures, can be selected using the operations and procedures key (OPS). In addition, information on costs per sector, days of incapacity for work and sick pay paid can be retrieved from the database. Data from insured persons from different service sectors can be combined in the database using an anonymized identification number, so that personal analyzes are possible over a longer period of time and across several sectors. The only possible interruptions in observation of insured persons in the database are a change of insured person or death.
A data set of around 4.9 million GKV insured people over a 6-year period (2014 – 2019) was used for our analyses.
The total costs were compared with the health insurance companies' income from the morbidity-oriented risk structure compensation (Morbi-RSA) in order to shed light on the income and expenditure side from the perspective of the statutory health insurance companies. The Morbi-RSA is intended to enable fair competition between statutory health insurance companies by balancing the financial consequences of the unequal distribution of expenditure risks, such as age, gender, place of residence and burden of illness, with a basic flat rate.
study population
In the first step, based on all insured persons in the database, insured persons with at least one confirmed outpatient or inpatient main or secondary diagnosis of diabetes mellitus (DM) (ICD-10-GM E10-E14) were identified in the period from January 1, 2014 to December 31, 2019. were selected that were consistently observable for at least four quarters. In the second step, insured people were selected from this group
i. at least one main inpatient DM diagnosis or
ii. had two inpatient secondary DM diagnoses/outpatient confirmed DM diagnoses (M2Q) or
iii. were prescribed at least one DM medication (see electronic appendix for ATC codes) in the same quarter of the confirmed outpatient DM diagnosis.
In the final step, patients with a confirmed outpatient or inpatient main or secondary diagnosis of a tumor disease (ICD-10-GM C00-C97 without C44) were excluded in order to avoid distortions due to the use of pain medication.
These patients were then stratified into two groups:
(a) Patients with DPN (ICD-10-GM G63.2) and
(b) Patients without DPN. Patients with DPN (group a) were again divided into patients with (1) s-DPN and (2) not s-DPN (see Fig. 1).
Patients with s-DPN were selected based on the diagnosis of “diabetic polyneuropathy” and the simultaneous prescription of defined pain therapies (e.g. gabapentin or pregabalin, full list see electronic appendix). Patients with a diagnosis of back pain (ICD-10-GM M45) or a knee endoprosthesis (OPS 5-82) in the same quarter of the first diabetes mellitus diagnosis were excluded and not assigned to any group (s-DPN or not s-DPN). in order to avoid potential distortions due to possible incorrect assignment of patients to s-DPN due to the pain medication related to the disease (back pain or knee endoprosthesis).
Analysis costs, sick pay, incapacity for work
The costs are shown as average costs per patient in euros from the perspective of the statutory health insurance companies. The total costs are also broken down into service sectors and compared with the health insurance companies' income from risk structure equalization (RSA) for the patient groups examined (DM with and without DPN, sDPN, not sDPN). For the analysis of sick days and sick pay, only employed insured persons were taken into account - pensioners and non-employed family insured persons were not included in these sub-analyses.
ethics
Since the work is based exclusively on secondary data, an ethics committee was not consulted. This corresponds to the recommendations of Guideline 1 of Good Practice for Secondary Data Analysis (Working Group on Collection and Use of Secondary Data (AGENS) et al. 2014).
Results
study population
From January 1, 2014 to December 31, 2019, the database counts a total of 4,874,745 GKV insured people (Step I), of which 550,825 had at least one diagnosis of diabetes mellitus coded and were observable for at least one year (Step II). After diagnostic validation (step III), 54,092 insured persons (9.8%) were excluded. After excluding the insured persons with a cancer diagnosis (step IV), the sample includes 434,615 DM patients without tumor disease (see Fig. 2), of which 20% (n=88,347) had DPN and 80% (n=346,268) did not have DPN were treated. Of the 88,347 patients with DPN, 18,111 patients (20%) were excluded due to the diagnoses of back pain and the knee endoprosthesis procedure. Subsequently, 30,552 patients (35%) with s-DPN and 39,684 patients (45%) with non-s-DPN were selected via pain medication prescriptions.
Characteristics of patients with and without diabetic polyneuropathy (DPN)
With an average age of 69 years, DM patients with DPN are around 5 years older than DM patients without DPN (64 years). Patients with s-DPN are around 3 years older than those with non-s-DPN (71 years versus 68 years). The proportion of male patients in both the group of patients with DPN (54%) and without DPN (51%) is slightly higher than that of female patients. However, 48% of patients with s-DPN are male and 61% of patients with non-s-DPN are male. Patients with DPN suffer from selected comorbidities more often than patients without DPN, whereas patients with s-DPN in particular show a higher burden of comorbidities. The characteristics of the patients with and without DPN as well as with s-DPN and with not s-DPN are shown in Table 1.
Cost
Patients with DPN incur total annual costs of an average of 13,433 euros - almost 48% of the costs arise from inpatient treatment. With total annual costs of 3,221 euros, patients without DPN incur around a quarter of the costs compared to patients with DPN. On average, the health insurance companies received a Morbi-RSA allocation of 9,614 euros annually for a patient with DPN - this corresponds to 72% of the actual costs incurred, whereas the Morbi-RSA allocations for patients without DPN are around the actual costs incurred 66% (see Fig. 3).
Patients with s-DPN incur around 22,266 euros annually, which is ~2.3 times higher than patients with non-s-DPN (9,727 euros) (see Fig. 4).
When looking at the total costs due to mental illnesses, patients with DPN have higher costs with an average of 71% than patients without DPN (2,197 euros, versus 1,285 euros). Of the patients with DPN, those with s-DPN in particular incur higher costs than patients with non-s-DPN (2,950 euros versus 1,876 euros).
Sick pay and incapacity for work
The average sums of sickness benefit payments over the period 2014-2019 for patients with DPN are around 27-99% significantly higher than for patients without DPN, with a continuous increase over time (see Fig. 5). If you take a closer look at the sickness benefit payments for patients with DPN, you can see that these are up to four times higher for s-DPN (year 2019: 2,710 euros vs. 620 euros) than for patients with non-s-DPN. The number of days of incapacity for work per year for patients with DPN is consistently higher than that for patients without DPN. Patients with DPN have an average of around 42 days of incapacity to work per year, while the average number of days of incapacity for work for patients without DPN is around 32. In particular, patients with s-DPN have a high number of days of incapacity for work (Ø: 47 days) compared to those with non-s-DPN (Ø: days 34).
discussion
summary of results
Based on a database with anonymized, age- and gender-adjusted billing data from approximately 4.9 million GKV insured people, 346,268 patients without DPN (80%) and 88,347 with DPN (20%) were identified from 434,615 DM patients.
Of the patients with DPN, 46% are female and one in three (n=30,552) patients suffer from s-DPN. The proportion of female patients is 52% for s-DPN and 39% for non-s-DPN. Compared to DM patients without a DPN, DM patients with a DPN cause around four times higher annual total costs (DPN: 13,433 euros versus without a DPN: 3,221 euros) and have higher sickness benefit payments (DPN: 1,502 euros versus without a DPN: 778 Euro) and days of incapacity for work (DPN: 39.6 versus without DPN: 31.9). Patients with s-DPN show total annual costs of 22,266 euros compared to 9,727 euros for patients with non-s-DPN. Sickness benefit payments (s-DPN: 2,710 euros versus not s-DPN: 620 euros) and days of incapacity for work (s-DPN: 35.5 versus not s-DPN: 33.7) also increase with the presence of s-DPN.
Classification of the results into existing literature
The proportion of DM patients with DPN (= 20%) determined in this work is comparable to existing literature. Köster et al. analyzed data from those insured by the AOK Hessen and described a prevalence of DPN in DM patients of 18.6% for the year 2010 (Köster et al. 2014). In comparison with international studies, differences in prevalence can be observed, which can be attributed to different study methods, diagnostics used, differentiation according to diabetes type, age, duration of diabetes or even ethnicity (Sun et al. 2020). Sun et al. evaluated data from 29 studies in different countries, published between 1993 and 2018, in a meta-analysis and reported a pooled prevalence in DM patients with DPN of 30%. The pooled prevalence in Europe (31.8%) was higher than in other regions, e.g. Asia (30.9%) or North America (22.6%) (Sun et al. 2020). We determined a prevalence of patients with s-DPN of 35%, which is comparatively high. Differently high prevalences in patients with s-DPN are reported, similar to Sun et al. described, influenced by various factors. Patients in particular who are older, female, have type 2 diabetes mellitus and have had diabetes for a longer period of time suffer from s-DPN (Abbott et al. 2011; Alleman et al. 2015; Ziegler 2020).
We also identified a higher proportion of women in patients with s-DPN (52%) compared to patients with non-s-DPN (39%). Similar results are reported by Kiyani et al. in their analysis of around 360,000 insured data from the USA, in which 53% of female patients showed s-DPN and 40% of female patients showed non-s-DPN (Kiyani et al. 2020). Furthermore, Geber et al. in their article that there is a connection between female gender and the development of s-DPN (Geber et al. 2010).
We analyzed average annual costs per DM patient without DPN of 3,221 euros. Köster et al. (Köster et al. 2014) reported total costs of 5,084 euros for a patient with diabetes mellitus in 2010, with 47% of the costs (2,391 euros) being attributable to the diabetes mellitus disease. Kähm et al. show data from those insured by the Techniker Krankenkasse (TK) and report average annual costs of 3,120 euros for a diabetes mellitus patient (70-79 years old) without complications (Kähm et al. 2018). We analyzed average annual costs per DM patient with DPN of 13,433 euros, i.e. four times higher costs for patients with DPN compared to patients without DPN. Köster et al. (Köster et al. 2014) show an increase in the 2,391 incremental costs due to diabetes mellitus disease by around 20% in the presence of DPN to 2,867 euros. The small increase in the study by Köster et al. is due, among other things, to the fact that patients with amputations were exclusively assigned to the “amputations” group - although we did not exclude the costs for amputations from the group of patients with DPN.
Results from several studies show that costs of DM patients with DPN are significantly higher than those of patients without neurological complications (Dworkin et al. 2011; Kiyani et al. 2020; Ritzwoller et al. 2009; Sadosky et al. 2015). Sadosky et al. were able to determine total direct medical costs of $6,632 for DM patients in the USA per year. In comparison, the costs for patients with DPN were 88% higher ($12,492) (Sadosky et al. 2015).
Furthermore, there are differences in the costs between s-DPN and non-s-DPN. We analyzed average annual costs per DM patient with s-DPN of 22,266 euros compared to DM patients with non-s-DPN of 9,727 euros, which is more than twice the cost. Similar results are reported by Kiyani et al. for the USA by analyzing data from around 360,000 insured people with diabetes mellitus. The average annual costs within 5 years of initial DM diagnosis (between 2010-2015) were 1.6 times higher for s-DPN than for non-s-DPN ($14,857 vs. $9,310) (Kiyani et al . 2020). Ritzwoller et al. report similar results for the USA. The average costs for a DM patient with non-s-DPN were estimated at $6,651 per year, while DM patients with s-DPN incurred an average annual cost of $14,062. A connection can also be observed with the intensity of pain. According to several studies, the costs increase as the severity of the pain increases (DiBonaventura et al. 2011; Kiyani et al. 2020; Sadosky et al. 2015).
Strengths and limitations
A sample representative of the population of those with statutory health insurance in Germany was used for the analysis. Since the database contains information from all service sectors, the total annual costs from the perspective of statutory health insurance in Germany could be identified for patients with and without DPN as well as with s-DPN and with non-s-DPN. The stratification of the total costs according to service sectors also made it possible to identify individual cost drivers.
The following limitations must be considered: Since there is no diagnosis of “s-DPN”, patients with s-DPN were identified via the diagnosis of DPN and the simultaneous prescription of pain medication. Patients with tumor diseases, with back pain and patients who received a knee endoprosthesis were excluded from the group of patients with s-DPN to avoid potential biases due to possible incorrect assignment of patients to s-DPN to avoid pain medication related to the illness (back pain or knee endoprosthesis). However, it cannot be conclusively ensured that patients with DPN received pain medication exclusively for the disease DPN and not for other underlying diseases. The costs are presented from the perspective of the service provider (statutory health insurance). Costs borne by those affected themselves are not taken into account. The study was purely descriptive. When comparing the costs, sick pay payments and days of incapacity for work of patients with and without DPN as well as with s-DPN and not s-DPN, it should be noted that these populations may differ with regard to other relevant factors, e.g. B. can influence the different cost structures (e.g. comorbidities).
Summary
The results of the study show that patients with DPN and especially with s-DPN cause significantly increased annual total costs from the perspective of the statutory health insurance compared to patients without DPN (DPN: 13,433 euros, s-DPN: 22,266 euros vs. without DPN: 3,221 euros). The costs are currently only partially covered by allocations from the Morbi-RSA (DPN: 72%, s-DPN: 53%). The increased comorbidity and disease burden with s-DPN, which can sometimes be the cause of the cost difference shown, also illustrates the burden on the sick patients. The need for therapeutic alternatives that serve to treat and relieve pain in s-DPN and thus increase the quality of life of patients seems high. <<
background
In Germany, 9-10% of the population is diagnosed with diabetes mellitus. Estimates put the number of DM patients at around 8.5 million, with an unreported number of at least 2 million (Tönnies et al. 2019). Sensorimotor diabetic polyneuropathy (DPN) is one of the most common complications that occur in connection with diabetes mellitus (Khdour 2020; Sadosky et al. 2015; Shillo et al. 2019). According to the American Diabetes Association, DPN is characterized by symptoms or signs of peripheral nerve dysfunction caused by the underlying disease diabetes mellitus (Shillo et al. 2019). The prevalence of DPN among diabetics is around 30% (Happich et al. 2008; Program for National Care Guidelines 2016; Sun et al. 2020; Ziegler 2020). While some of the affected patients show asymptomatic disease, 8-34% suffer from neuropathic pain (Shillo et al. 2019; Ziegler 2020). Patients with s-DPN have burning pain and discomfort as well as unpleasant cold or heat paresthesia, especially in the feet and lower legs (Hilz et al. 2000; Sadosky et al. 2015; Shillo et al. 2019; Ziegler 2020), which increase in perception at night. Disturbances in the sense of touch, position and vibration are among the symptoms of non-s-DPN (Hilz et al. 2000). Risk factors for the development of DPN include older age, increasing duration of diabetes, inadequate diabetes control, excessive alcohol consumption, smoking and lack of physical activity (Happich et al. 2008; Ziegler 2020). DPN, especially the non-painful form, increases the risk of diabetic foot ulcers due to the lack of sensation with the risk of subsequent amputation (Happich et al. 2008). DM patients have a 10 to 22 times higher risk of amputation compared to people without diabetes mellitus (Happich et al. 2008; Khdour 2020; Program for National Care Guidelines 2016; Ziegler 2020). Patients with chronic pain suffer from significant restrictions in their daily lives, which often leads to psychological comorbidities such as anxiety, depression, sleep and mood disorders (Khdour 2020; Shillo et al. 2019). Overall, DPN is associated with impaired quality of life, an increased risk of mortality and significant morbidity (Happich et al. 2008; Khdour 2020; Program for National Care Guidelines 2016; Sadosky et al. 2015; Shillo et al. 2019; Ziegler 2020).
