Development and testing of diagnostic algorithms to identify patients with acromegaly in southern Italy claims databases

data sources

This Italian retrospective, population-based study was conducted from January 2015 to December 2018, using data extracted from fully anonymous Administrative Records Linkage databases of the Caserta Local Health Unit (LHU), with an average of 1,060,904 inhabitants. Furthermore, data from fully anonymous administrative health databases for the territory of Sicily from January 2011 to December 2018, with an average of 5,031,655 populations, were used as a test dataset to evaluate the performance of the proposed algorithms and substantiate the estimates of the identified megalimb cases. In Caserta LHU. Both databases contain demographic and medical data collected through services provided by the Italian National Health Service (NHS). They include information on the demographics of residents in each pool area, outpatient pharmacy claims, hospital discharges, co-payment waivers, referrals for outpatient diagnostic tests and a database of specialist visits. The content and context of the NHS claims databases have been described in detail elsewhere19. The claims databases from Caserta LHU also include the Electronic Therapeutic Plans (ETP) database. In Italy, ETPs are filled in directly by professionals, who provide information on the exact brand name, the number of packets dispensed, and an indication for use. Dispensed drugs were coded using the Anatomical Therapeutic Classification System (ATC) and the Italian Marketing Authorization Act (AIC), while comorbidities and indications for use were coded by the 9th revision of the International Classification of Diseases – Clinical Modification (ICD-9) cm.

Algorithm definition

To detect the presence of acromegaly in the target population, four different algorithms have been proposed and developed based on a systematic review of published articles from megacromegaly epidemiology using claims database-based algorithms.2. Specifically, each algorithm was developed using a combination of acromegaly-related ICD-9 CM diagnostic codes (253.0), ICD-9 CM surgical procedure codes (07.6x and 92.3x), and common payment exemption codes (001 and 253.0). , specialist visits, lab exams, and pharmacy claims for somatostatin analogs (ATC: H01CB02, H01CB03, H01CB05) or pegvisomant (ATC: H01AX01). Algorithm 1 was developed by Caputo et al. in 20184“Algorithm 2” is from a review of “Algorithm 1,” which included more code related to edge amplification, while “Algorithm 3” and “Algorithm 4” were developed through different sets of code related to edge size used in the publication.3,4,21And the2223. For each algorithm, as suggested by Caputo et al.4, pharmacy claims for drugs approved for the treatment of acromegaly were not considered if (i) patients received fewer than three separate prescriptions for acromegaly (occasional drug users) during the observation period; (2) The drugs were not long-acting (LAR) formulations; (3) Patients taking octreotide or lanreotide have had at least one hospital admission with a diagnosis that, as stated in the summary of product characteristics, is one of the indications for use of these drugs. [i.e., malignant neoplasms (ICD-9 CM: 140–209, 230–239), liver disorders (ICD-9 CM: 570–573), gastrointestinal bleeding (ICD-9 CM: 578), esophageal varices (ICD-9 CM: 42), Cushing’s disease (ICD-9 CM: 255; 255.0)]; (4) Patients had the Cushing’s Disease Co-Payment Waiver Code (Code: 032). The inclusion and exclusion criteria are reported in Table 1.

Table 1 includes inclusion and exclusion criteria for each proposed algorithm for identifying cases of acromegaly. * Italian coding system.

Definition of the gold standard cohort

In the Caserta LHU, gold-standard cases were defined as subjects who had at least one entry in the ETP database with at least one ICD-9 CM code for acromegaly during the study period. Non-gold-standard cases were defined as all remaining subjects (that is, registered in the Caserta LHU without ICD-9 CM codes for acromegaly in the ETP database).

In the Sicily region, gold standard cases were defined as patients with a confirmed diagnosis of acromegaly in the endocrine unit of the University Hospital of Messina, a province located in the northeastern part of Sicily, during the study period. Data on non-cases is not available.

statistical analysis

For each proposed algorithm, accuracy was assessed by measures of sensitivity (SE) and specificity (SP), along with 95% confidence intervals (CIs) computed using the exact Clopper-Pearson binomial ratio method. Furthermore, the Youden Index (i.e. a summary statistic that balances both SE and SP)24 It was computed and the algorithm with the highest accuracy (i.e. Yoden index) was considered to be the preferred algorithm compared to the other algorithm. Moreover, positive predictive value (PPV) and negative predictive value (NPV) along with 95% CIs were also estimated to assess the accuracy of the algorithm. To evaluate the overlapping number of acromegaly states detected by each optical coding algorithm with respect to different data sources, Venn diagrams were produced.

The four proposed algorithms were also applied to patients diagnosed with acromegaly in the endocrine unit of Messina University Hospital by connecting a probabilistic log and the proportion of true positives among these cases was compared with the achieved SE for each algorithm. In addition, to provide an estimate of the percentage of acromegaly cases in the Sicilian population, the algorithm with the highest discriminatory power was implemented only in the administrative data of a Sicilian region (test dataset).

Finally, for the algorithm with the highest Youden index, a network plot was produced to show the number of identified acromegaly patients in the Caserta LHU and the Sicily region, respectively, to demonstrate the temporal frequency of any acromegaly code identified in each claims database. In particular, for each claims database, all possible pathways by which subjects were identified over time were represented. Each pathway consists of the chronology in which each patient’s acromegaly symbol occurred. This can be considered as a proxy for the patient’s journey25.

All statistical analyzes were performed using the R Foundation for Statistical Computing (R Development Core Team 2008, Vienna, Austria, version: 4.0.3, packages: caret And the PropCIs).

Ethics approval

Analyzes were performed in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki Declaration and its subsequent amendments. This study was approved by the Ethics Committee of Integrated university hospital Verona (Protocol No. 55986, September 27, 2021).

Agree to share

Informed consent was obtained from all individual patients from the endocrinology unit of Messina University Hospital included in the study.