Abstract
Little is known about the diagnostic methods currently used in routine care for patients with mild cognitive impairment (PwMCI). We estimated the frequency of diagnostic procedures in incident PwMCI compared to incident patients with dementia (PwD) in 2016-2017. The study is based on the Disease Analyzer database. After matching by age and sex, 4,700 PwMCI and 4,700 PwD were available. The diagnostic procedures were identified on the basis of the related medical fee schedule items. All diagnostic procedures were used more frequently in PwMCI than in PwD. The drafting of a practice-oriented MCI guideline is an important task for the future.
INTRODUCTION
Care for patients with dementia is facing major challenges. In the future, such care will focus on the early stages of the disease [1 –5]. In particular, the inclusion of patients with mild cognitive impairment (PwMCI) and the subgroup of patients with prodromal Alzheimer’s disease (MCI due to Alzheimer’s disease) identified on the basis of biomarkers will more than double the number of patients to be treated. The current prevalence of dementia in Germany is estimated at around 1.7 million patients [6]. A current forecast of the RAND Corporation report estimates that in 2019, 1.7 million of a possible 3.7 million MCI-positive patients could benefit from biomarker-based diagnostics [7]. This would mean that the ratio of patients with dementia (PwD) to PwMCI would be about one to two. It has recently been reported that while the number of diagnosed PwMCI in routine care is increasing, it is still far below the expected prevalence [8].
ICD-10 code F.06.7 is available as a code for documenting the diagnosis of MCI. A medical history and the collection of clinical findings are sufficient to fulfil the required diagnostic criteria [9]. When diagnosing MCI in the context of Alzheimer’s disease, however, examinations that can identify biomarkers are required [5]. In addition to neuropsychological testing, these include the examination of cerebrospinal fluid, cerebral CT imaging or MRI and PET examinations. The extent to which these examinations are actually performed or ordered by general practitioners (GPs) and neuropsychiatrists (NPs) treating statutorily insured patients is currently unknown. The aim of this study was to clarify the extent to which neuropsychological tests, blood sampling, cerebrospinal fluid punctures, and radiological examinations were performed or ordered in PwMCI compared to a control group consisting of PwD in between January 1, 2016 and December 31, 2017.
METHODS
Data source
The present retrospective study used data from the nationwide Disease Analyzer database (IQVIA). The data contained in this database originate from a nationwide sample of general and specialized practices [10]. The quality of the information is reviewed regularly by IQVIA, and the representativeness of the Disease Analyzer database for German primary care practices has already been confirmed by previous studies [10].
Study population
Those patients for whom an MCI diagnosis had first been documented between January 1, 2016 and December 31, 2017 were selected from a sample of more than 4 million patients from 1,104 GP practices and 138 NP practices. In the six months prior to and the six months after the diagnosis, the following services were assessed using the medical fee schedule items (GOP) specified in the German Uniform Assessment Standard (EBM) (https://www.kbv.de/html/ebm.php): blood sampling (GOPs of selected laboratory values (including fasting plasma glucose, HbA1c, leukocytes, erythrocytes, hemoglobin, hematocrit, creatinine, bilirubin, cholesterol, and others), neuropsychological testing (GOP 03242, 03360, 16340, 212340, 35300/35600, 35301/35601), lumbar puncture (GOP 02342), and referrals for radiological evaluations. Controls with dementia were assigned 1:1 to PwMCI based on age, sex, and physician’s practice. This matching procedure reduced the number of GP practices from N = 1,044 to N = 825 and the number of NP practices from N = 137 to N = 132 (cf. Table 1).
Total sample of patients with incident dementia and MCI (2016-17) and 1:1 matching by age group and sex
Statistical analyses
The main outcome of the study was the difference between proportion of PmMCI and PwD with at least one blood sampling or referral to a laboratory physician, neuropsychological testing, referral for radiological evaluations and lumbar puncture. Descriptive analyses were obtained for demographic variables (age and sex), and mean±SD were calculated for age. χ 2-tests were used to test the difference between proportions of PmMCI and PmD. A p-value of <0.05 was considered statistically significant. The analyses were carried out using SAS 9.4.
RESULTS
A total of 30,104 patients in GP practices (N = 1,044) received an initial dementia diagnosis and 2,316 patients were diagnosed with MCI in the study period between January 1, 2016 and December 31, 2017. In NP practices (N = 137), the figures were 14,035 and 2,384, respectively. The dementia and MCI samples differed significantly in age and sex in both GP and NP practices. The mean age of PwD was 81.8 years (GP) and 81.3 years (NP), while for PwMCI it was 75.1 years (GP) and 70.7 years (NP). The share of women in GP practices was 63.4% for PwD and 57.8% for PwMCI; in NP practices the figures were 63.1% and 51.6%, respectively. PwD were matched 1:1 to PwMCI (cf. Table 1)
The tests administered in PwD and PwMCI in GP and NP practices are listed in Table 2. Tests were administered in 34% of PwD and 63% of PwMCI in GP practices and in 35.8% and 54.2%, respectively, in NP practices; laboratory values were determined for 29.9% of PwD and 65.2% of PwMCI in GP practices and in 3% and 8.4%, respectively, in NP practices. A total of 11.1% of PwD and 27.5% of PwMCI in GP practices received referrals to radiologists; the figures were 21.7% and 40.9%, respectively, in NP practices. Cerebrospinal fluid punctures were only performed in NP practices (6 PwD: 0.3% versus 13 PwMCI: 0.5%) (cf. Table 2).
Comparison of dementia versus MCI with regard to test procedure, laboratory, radiology, and cerebrospinal fluid puncture after 1:1 matching
* χ 2-test, p < 0.05.
Prior to 1:1 matching, tests, laboratory examinations, and radiological examination were performed or ordered for 48.8%, 39.9%, and 12.6% of PwD in GP practices respectively. In NP practices, the numbers were 31.8%, 2.3%, and 17.6% respectively. Cerebrospinal fluid punctures were only performed in NP practices (0.3%) (cf. Table 2).
DISCUSSION
PwMCI in GP and NP practices were examined using the same diagnostic procedures as PwD. Cerebrospinal fluid punctures, which were only performed by NPs, were the only exception. PwMCI received all four diagnostic tests (testing, laboratory, radiology, cerebrospinal fluid) more frequently than PwD. It can be assumed that PwMCI were examined more closely compared to PwD due to the greater diagnostic effort.
We are not aware of any national or international studies in routine care research that reported the frequency of diagnostic procedures used in MCI patients. In our discussion, we therefore only consider care studies that report such frequencies in diagnosing PwD. All examination procedures have been available in specialized memory clinics, especially at universities, for over 10 years and are part of the standard diagnostic procedures in these settings [11].
With regard to the frequency of neuropsychological tests administered in PwD reported here for 2016-2017 (34% in GP and 35.8% in NP practices), no change on the data from 2009 can be observed. Schulz et al. reported a rate of 34% for the overall number of PwD [12]. In 2012, Thyrian et al. reported that neuropsychological screening instruments were used in nearly 80% of Mecklenburg GP practices, albeit without giving any indication of the frequency of their use [13]. The psychological testing rates reported here for PwMCI (63% in GP and 54.2% in NP practices) are therefore higher than those reported for PwD in older studies.
The laboratory test rate among PwD in a complete survey in 2009 was 79.2% [12] and thus well above the significantly lower rates reported here. This is mainly due to the fact that most GP and NP are part of collaborative laboratories whose data were not available for the present study. Van den Bussche et al. found that just 2% of patients received dementia-specific laboratory diagnostics (e.g., TSH and vitamin B12) in 2004 [14]. In an intervention study including 129 GP and 390 patients, Donath et al. found that more than 95% of PwD received laboratory testing [15]. For the purposes of the present study, it can at least be assumed that those practices that do not bill their laboratory services through collaborative laboratories perform laboratory examinations more frequently in PwMCI than in PwD.
The share of PwMCI referred for radiological examinations is surprisingly high compared to the complete survey conducted in 2009 cited above, which reported that 18% of PwD were examined using MRI or CT technology. However, higher rates than in routine care were also reported for PwD in intervention studies as part of study practices. Thus, 45% of PwD in the study by Donath et al. [14] and 85% of PwD in a specialist sample in Mecklenburg-Vorpommern [16] underwent imaging. The higher rates of imaging examinations in intervention studies may be subject to selection bias, but this cannot be assessed here.
The rate of cerebrospinal fluid examinations among PwMCI has not yet been investigated in routine care. These examinations are usually specific to memory clinics; they are not performed at all in GP practices and only in exceptional cases in NP practices. Only minor differences were observed between PwMCI and PwD.
Similar to German studies [17, 18], international survey studies, in which GPs were asked about their attitudes toward the diagnosis of MCI, indicated that the acceptance of diagnostic measures in MCI and in the very early stages of dementia has increased [19 –22]. Various efforts are also being made to improve the timely diagnosis of dementia and to ensure that any barriers to diagnosis are taken adequately into consideration [23, 24]. However, few studies have investigated how PwMCI have been diagnosed in routine care. A significant increase in MCI diagnoses in GP and NP practices has been reported in recent years. The MCI prevalences in GP and NP practices tripled and doubled, respectively, between 2007 and 2017 [8], yet MCI was still rarely diagnosed. The coded MCI diagnosis prevalence rate of 10%, which is very low compared to the true prevalence rate, suggests that examination routines for PwMCI had not yet developed in the practices as of 2017. It is therefore all the more astonishing that the patients in the present sample were examined more extensively and possibly more thoroughly than PwD.
The following methodological limitations have to be considered. The concept of MCI as listed in the ICD-10 is unspecific in routine care and can cover a broad range of disorders. Not every MCI patient with the ICD-10 diagnosis (F 06.7) is necessarily a patient who has MCI due to Alzheimer’s disease, according to the DSM-5. Various other conditions may hide behind this diagnosis. It is conceivable that MCI diagnoses may also be made without a psychological testing [25]. The type of psychological tests administered was not specified. The GOP of the EBM can indicate complex tests as well as brief screenings. The information available regarding laboratory tests did not include any tests performed in collaborative laboratories. The frequency of radiological examinations was only mapped on the basis of referrals to radiology services issued by the respective GP or NP practice. Since we were unable to check the exact content of the referral, it was not possible to determine whether the brain or another anatomical region was to be examined, nor could the examination method be specified.
Cerebrospinal fluid punctures are only reimbursed when performed by NP. No information was available about the exact cerebrospinal fluid parameters examined. Only statutorily insured patients were examined, as information on the diagnostic procedures included was available for these patients in the form of uniform medical fee schedule items. In the case of privately insured patients, these services are paid for via a different reimbursement system. However, these patients only account for less than 10% of the total. Due to the matching method, a part of PmMCI were excluded from the analyses. This reduced the total patient numbers. Data from hospitals was lacking. When patients were hospitalized and blood sampling or neuropsychological testing were performed in the hospital, this information was not available.
The strength of this study lies in the fact that it provides the first description of examination procedures used in routine care—blood sampling, testing, cerebral imaging, lumbar puncture—and includes a large number of PwMCI and a representative sample of GP and NP practices. This resulted in the first available indications regarding the diagnostic habits found in a real-world setting. We were able to show that PwMCI were examined more extensively than PwD. However, in view of the low prevalence of coded MCI diagnoses, there is considerable room for improvement of the diagnostic procedures. The drafting of diagnostic guidelines and treatment guidelines for PwMCI is therefore desirable [5].
DISCLOSURE STATEMENT
Authors’ disclosures available online (https://www.j-alz.com/manuscript-disclosures/19-0012r1).