Abstract

Objective. To calculate the full cost of diagnostic pathology tests for Non-Small Cell Lung Cancer (NSCLC) across four Italian Pathology Units.
Methods. Pathology Units were located in private (2) and public (2) hospitals distributed across the Italian territory (North: 2; Centre: 1; South: 1). Pathologists provided via questionnaire data on tests on NSCLC samples along with the identification and quantification of the necessary healthcare resources (diagnostic technologies, laboratory instruments and personnel). Resources were valued according to hospital-specific unit, yearly and hourly costs (disposables; technologies; professional clusters).
Results. The full cost per NSCLC tissue sample included histopathological immunophenotypic and required molecular analysis. Overall, it reached € 659.77 and it was mainly composed of direct costs (77.69%). The processing of a NSCLC tissue sample was labour intensive, as a relevant share of the full cost (44.98%) was actually due to personnel costs, with laboratory technicians, biologists and pathologist driving this finding (17.09%,12.43% and 10.81%, respectively).
Conclusions. The results of this research can facilitate the negotiation of new dedicated tariffs for NSCLC sample processing with the national or local third party-payers.

Introduction

In Italy, in 2020, there had been about 41,000 cases of lung cancer (LC) which remains the second most frequent cause of malignancy in men and the third in women. Compared to recent years that take into account the population cancer registries from 2008 to 2016, the incidence trend decreased by 1.7% in men and increased by 3.4% in women. In terms of mortality, LC represents the first cause of death in men and the second in women among all cancers ^1,2.

The histopathological classification includes four main histotypes: squamous cell carcinoma; adenocarcinoma; large cell carcinoma; small cell carcinoma ^2,3.

The diagnosis can be performed on the basis of a careful evaluation of conventional morphological criteria on hematoxylin-eosin or specific stains (e.g., May-Grunwald-Giemsa or Papanicolaou) for cytological samples. Immunohistochemical (IHC) investigations are fundamental for the precise definition of the poorly differentiated or not otherwise specified Non-Small Cell Lung Cancer (NSCLC) types ³.

The IHC investigations can be applied both to biopsy samples fixed in formalin and to cytological preparations. The preparation of cell-blocks starting from cytological material on effusion or aspiration is of great practical value, especially in consideration of the possibility of analyzing through IHC some of the predictive biomarkers of response to medical therapy, such as the Programmed Death-Ligand 1 (PD-L1) ⁴.

Lately, the use of molecular investigations like next generation sequencing (NGS) has grown in specialized centers with highly experienced pathologists. The molecular characterization of lung tumors is a fundamental element of the patient’s diagnosis and treatment process, in light of the possibility of recommending molecular targeted treatments in populations selected for the presence or expression of a certain marker. In this regard, in all patients with NSCLC in stage IIIB-IIIC (not candidates for loco-regional treatments), and IV, it is recommended to integrate the morphological diagnosis with the molecular characterization of the following genes: EGFR (Epidermal Growth Factor Receptor); BRAF (B-Raf proto-oncogene) mutations; ALK (Anaplastic Lymphoma Kinase); ROS-1 (Proto-oncogene tyrosine-protein kinase ROS); RET gene rearrangements; NTRK (Neurotrophic Tyrosine Receptor Kinase) translocations ^2,5.

In addition, there are molecular alterations for which there are currently no drugs approved and reimbursed in Italy, but available only in the context of clinical trials and compassionate use programs or early access programs, such as: MET exon 14, KRAS exon 2 (G12C), PI3KCA, PTEN and activating HER-2 gene mutations or FGFR1 and PDGFR amplifications ^4-6.

The current recommendations developed by the collaboration of SIAPeC (Società Italiana di Anatomia Patologica e Citologia diagnostica) and AIOM (Associazione Italiana di Oncologia Medica) suggest the concomitant evaluation of gene mutations, of ALK, ROS1, NTRK 1-3 rearrangements, and the PD-L1 expression profile in tumor cells for the choice of the best therapeutic strategy mostly in advanced stage NSCLC patients ⁷.

This paper reports on an empirical cost description ⁸ of the diagnostic pathology tests performed on NSCLC samples across four Italian Pathology Units at the forefront of NSCLC diagnosis.

Materials and methods

DATA COLLECTION

A questionnaire was developed and sent out by e-mail to a convenience sample ⁹ of six pathologists between October-November 2020 to receive their feedbacks and validation.

Pathologists practiced in four Pathology Units located in private (2) and public (2) hospitals distributed across the Italian territory (North: 2; Centre: 1; South: 1). Four out of six pathologists practiced in the same Pathology Unit (FB, MA), or in the Pathology Department of the same hospital (PC, MGP); the other two pathologists (MB, PG) worked in different Pathology Units.

All the Pathology Units are endorsed as healthcare service providers by the Italian National Health Service (INHS).

Upon questionnaire approval, experts were requested to provide the following quantitative data: overall number of patients referred to their Pathology Unit; overall volume of tissue samples, blocks, and slides along with the fraction of these specimens related to NSCLC diagnosis (Tab. I).

Data on tests for NSCLC along with the identification and quantification of the necessary healthcare resources (diagnostic technologies, laboratory instruments and personnel) were also detailed in the questionnaire.

When needed, follow-up teleconferences and on-site visits were scheduled with the experts to fix inconsistencies.

All data deliberately refer to 2019, the year before the COVID 19 pandemic outbreak. This methodological choice seems justified in the light of the possible standard activity reduction in all the non-COVID 19-dedicated hospital wards in 2020 that, if considered as a yardstick, might have provided an unreliable representation of the average number of patients and tests processed by each Pathology Unit included in this study.

As the present research did not require patients’ enrollment, no Ethics Committee approval of the questionnaire was required by the current Italian legislation ¹⁰.

COST

A cost description ⁸ (i.e., the valuation of the resources needed to perform a given healthcare service) of two cost objects ¹¹ (i.e., two items of interest for which the cost is calculated, that is NSCLC tissue sample and NSCLC patient in this research) was carried out. The cost description followed the Pathology Unit viewpoint ⁸. Therefore, only costs borne by the Pathology Units that participated in this research were considered.

The cost of a NSCLC patient was obtained by multiplying the cost of a NSCLC tissue sample by the NSCLC tissue samples/NSCLC patient ratio (1.720).

The cost description adopted a full cost approach ¹¹.

The full cost approach implies three types of costs: direct and indirect costs of the Pathology Unit, and overheads ¹¹.

Direct costs value all the resources that can be specifically referred to a single cost object (e.g., the number of minutes needed by a biologist to perform a KRAS test on a NSCLC tissue sample multiplied by her/his gross hourly wage).

Indirect costs include the value of resources that are common to many cost objects (e.g., a microscope is used by pathologists for diverse analyses in addition to the confirmation of a suspected NSCLC diagnosis). Therefore, determining the share of their amount to be assigned to a given cost object is physically unfeasible and/or too expensive.

Moreover, oftentimes indirect costs include resources that are expected to contribute to the activity of the Pathology Unit well after the year of purchasing. Therefore, their purchase cost is allocated over their useful life according to different accounting procedures (e.g., straight line approach, in which the overall cost of the microscope is divided into equal shares of cost per year of its useful life, which is estimated to be 10 years in this research) that go under the name of depreciation ⁸.

A share of the indirect costs was imputed (i.e., assigned) to the NSCLC tissue sample via specific allocation bases (e.g., the annual depreciation of the microscope divided by the overall usages made by the pathologists during that year multiplied by the number of slides stained for a specific NSCLC sample).

Overheads represent fair shares of direct and indirect costs totaled by technostructure (e.g., administration; accounting) and support staff (e.g., cleaning; building maintenance) of the hospital ¹².

As technostructure and support staff contribute to the activity of hospital wards and services, their costs were imputed pro-quota to Pathology Units. The shares of overheads imputed to Pathology Units were estimated by multiplying the sum of direct and indirect costs of a NSCLC tissue sample and NSCLC patient by 125%.

Resources were valued according to hospital-specific unit costs (technologies; disposables) and personnel costs (gross hourly cost for different professional clusters) (Tab. II).

The full cost of a NSCLC tissue sample was broken down into the four main steps its production is composed of: Step 1 - Pre-analysis; Step 2 - Analysis - Staining; Step 3 - Analysis - Molecular; Step 4 - Finalization and medical reporting. Whenever necessary each step was detailed in substeps (Tabs. SI-SV).

The full cost of a NSCLC diagnostic characterisation was obtained by averaging over the cost of all the tests necessary to confirm a NSCLC diagnosis.

As anticipated, costs, expressed in Euros (€ ), date back to 2019.

Results

The full cost per NSCLC sample and per patient was € 659.77 and € 1134.61, respectively, and was mainly composed of direct costs (77.69%) (Tab. III).

Step 3 (Analysis – Molecular) was the most expensive step of the NSCLC tissue sample (€ 239.13), whereas Step 1 (Pre-Analysis) was the cheapest one (€ 71.88). No share of indirect costs was imputed to Step 3 (Analysis - Molecular).

The processing of a NSCLC tissue sample was labour intensive, as a relevant share of the full cost (44.98%) was actually due to personnel costs, with laboratory technicians, biologists and pathologist driving this finding (17.09%,12.43% and 10.81%, respectively).

As far as indirect costs are concerned, IHC stainers were the most remarkable cost item (0.91% of the full cost).

The same percentages hold for the full cost per NSCLC patient, as it was calculated by multiplying the cost per NSCLC tissue sample by a 1.720 constant.

The very same trend has been observed by analyzing in detail each one of the four main steps the production of a NSCLC tissue sample was divided into.

The full cost was steadily led by direct costs, with percentages ranging between 71.18% (Step 1 - Pre-analysis) and 80.00% (Step 3 - Analysis - Molecular), with personnel cost still playing a relevant role, varying from 15.07% (Step 2 - Analysis - Staining) to 78.93% (Step 4 - Finalization and medical reporting).

The indirect costs/full cost ratio ranged between 0.00% (Step 3 - Analysis - Molecular) and 8.82% (Step 1 - Pre-analysis).

DISCUSSION

While the INHS introduced the so-called F file to make expensive hospital drugs available on the Italian territory without stringent financial constraints ¹³, a similar mechanism aimed at facilitating the widespread adoption of advanced healthcare technologies for the diagnosis of NSCLS histotypes via an easy to apply, fair reimbursement scheme does not exist in Italy, yet.

In fact, there is always a lag between the pace at which healthcare science and technologies progress and the INHS decision to reimburse the healthcare services that can benefit from them.

A tariff-based system for reimbursing outpatient healthcare services provided by both public and INHS-endorsed private hospitals, was adopted by the INHS in the early 1990s and was last updated in 2013 ¹⁴.

For the sake of precision, a more recent outpatient healthcare services’ handbook does exist (named Essential Levels of Assistance 2017). However, at the time of writing, the related tariffs proposed by the Ministry of Health in 2022 have not been approved by the State-Regions Conference, yet.

Interestingly, Italian regions are free to adopt their own tariff-based system (that can well include healthcare services that are not reported in the INHS outpatient healthcare services’ handbook, with the provision that regional autonomy comes with the local effort to fund them without any support from the central government) or stick with the national one.

At both national and local levels, a tariff-based reimbursement system pursues two different but related goals. First, setting, without any mandatory negotiation with the providers, the price at which the third-party payer, acting as monopsonist (i.e., the agent that purchases the highest share of the inpatient and outpatient healthcare services produced by public and INHS-endorsed private healthcare organizations) is willing to pay for a given healthcare service. Second, allowing the third-party payer to curb the volume of overprescribed healthcare procedures (usually the most profitable ones for the healthcare organizations) by reducing their tariffs, whereas increasing the reimbursements for less costly alternatives that are expected to have the same diagnostic accuracy or effectiveness on patient’s health.

Consequently, the difference between costs and tariffs becomes crystal-clear: while costs express the economic value of the resources needed to perform a given healthcare procedure when displaced from the best available alternative (cost-opportunity principle), tariffs are an administrative tool aimed at managing the healthcare system procurement activity ¹⁵.

Therefore, theoretically speaking, a well-conceived tariff-based reimbursement system should impose on the third-party payer frequent assessments of the full cost actually borne by healthcare organizations to provide patients with a given healthcare service. This would help to align the tariffs to the value of the resources actually consumed by the healthcare organizations.

In addition, a systematic horizon scanning could help the third-party payer to calculate new tariffs for advanced healthcare procedures for a given disease to be adopted by healthcare organizations consistent with the international guidelines.

Is this the case for the NSCLC sample processing? Unfortunately, the answer is not encouraging.

Skimming through the 2013 INHS outpatient healthcare services’ handbook, only in situ FISH hybridization (four different tariffs, coded 91.37.2-91.37.5, ranging from € 150.29 to € 342.87) and molecular probe hybridization (one tariff only, coded 91.37.1; € 81.60) are reported ¹⁴.

It is clear that the generalized lack of specific INHS tariffs aimed at reimbursing NSCLC diagnostic tests disincentives their widespread adoption and creates sustainability issues for the hospitals patients with a suspected NSCLC are referred to. As a consequence, a limited availability of these molecular tests reduces the likelihood of a homogeneous diagnosis across the Italian territory. In turn, this disparity may hamper the equal access to the most appropriate therapy for NSCLC patients.

The main finding of our study is the detailed breakdown of the real full cost borne by a convenience sample ⁹ of four Pathology Units at the forefront of the diagnosis and molecular characterisation of NSCLC in Italy.

However, the most striking result of the empirical research detailed in the previous paragraphs is the extreme relevance of the personnel contribution.

The relationship between NSCLC tissue sample processing and labour intensity in the Italian setting was previously highlighted ^16,17.

A 5-year budget impact analysis proved the Ventana ALK (D5F3) CDx assay to reduce the oncologists’ work load in detecting ALK positive NSCLC patients eligible to crizotinib ¹⁶.

In a similar research performed on a sample of 1461 NSCLC patients followed up at 5 Italian oncology units, the scenario that assumed the widest adoption of next-generation sequencing resulted in an overall average personnel time of 1975 hours per center totaled by technicians, biologists and pathologists involved in molecular analysis ¹⁷. However, from the data reported in the article the conversion of personnel time in costs was not feasible.

Actually, it takes years of formal education and on-the-job training ¹² to become a proficient and autonomous pathologist dealing with NSCLC sample processing, and, likewise other sectors of medical sciences (e.g., diagnostic imaging), technology cannot completely replace human ability, in terms of reliability of the final outcome of the healthcare procedure, which, in this case, has a substantive bearing on the subsequent appropriate target therapy ⁷.

The main limitation of this study relates to the small number of pathological anatomy centers investigated and their top-level qualification and expertise in the diagnostic pathology of NSCLC. Therefore, the external validity of our results (especially in terms of direct costs/full cost ratio) should be proved by future, larger empirical studies carried out on a larger sample of Italian Pathology Units.

Conclusions

Priority setting in healthcare should be made on the grounds of real-world evidence.

We are confident that, despite its limitation, the approach adopted in this research can facilitate the negotiation of new dedicated tariffs for NSCLC sample processing with the national or local third party-payers. Hopefully, this study will pave the way to future, empirical cost descriptions aimed at addressing the same topic for other human cancers.

CONFLICTS OF INTEREST

SC and FDP are employees of Roche Diagnostics S.p.A. Monza, Italy. The other authors declare no conflict of interest with this research.

FUNDING

This study was funded by an unrestricted grant from Roche Diagnostics S.p.A., Monza, Italy, that contributed to study design, data acquisition and analysis.

ETHICAL CONSIDERATION

Not applicable.

AUTHORS’ CONTRIBUTION

Study design: GF, MB, FB, PC, PG, AM, SC, FDP, MGP. Data acquisition; MB, FB, PC, PG, AM, SC, FDP, MGP; Data analysis: CL, GF, MB, FB, PC, PG, AM, SC, FDP, MGP. Data interpretation: CL, GF, MB, FB, PC, PG, AM, SC, FDP, MGP; Manuscript preparation: CL,GF. Manuscript review: CL, GF, MB, FB, PC, PG, AM, SC, FDP, MG.

Figures and tables

References

1. Passiglia F, Calandri M, Guerrera F. Lung cancer in Italy. J Thorac Oncol. 2019; 14:2046-2052. DOI
2. WHO Classification of Tumours Editorial board (Editors). Thoracic Tumours. WHO Classification of Tumours. IARC: Lyon; 2021.
3. Nicholson AG, Tsao MS, Beasley MB. The 2021 WHO classification of lung tumors: impact of advances since 2015. J Thorac Oncol. 2022; 17:362-387. DOI
4. Mino-Kenudson M, Schalper K, Cooper W, IASLC Pathology Committee. Predictive biomarkers for immunotherapy in lung cancer: perspective from the International Association for the Study of Lung Cancer Pathology Committee. J Thorac Oncol. 2022; 17:1335-1354. DOI
5. Lindeman NI, Cagle PT, Aisner DL. Updated molecular testing guideline for the selection of lung cancer patients for treatment with targeted tyrosine kinase inhibitors: guideline from the College of American Pathologists, the International Association for the Study of Lung Cancer, and the Association for Molecular Pathology. J Thorac Oncol. 2018; 13:323-358. DOI
6. Angerilli V, Galuppini F, Pagni F. The role of the pathologist in the next-generation era of tumor molecular characterization. Diagnostics (Basel). 2021; 11:339. DOI
7. Associazione Italiana di Oncologia Medica (AIOM). Linee guida per le neoplasie del polmone. AIOM: Milan; 2021. Publisher Full Text
8. Drummond MF, Sculpher MJ, Claxton K. Methods for the economic evaluation of health care programmes. Oxford University Press: Oxford; 2015.
9. Lohr SL. Sampling. Design and Analysis. Chapman and Hall/CRC: Boca Raton; 2021.
10. Ministero della Salute. Decreto 8 febbraio 2013. Criteri per la composizione e il funzionamento dei comitati etici. (13A03474). Gazzetta Ufficiale della Repubblica Italiana. 2013; 96
11. Atkinson AA, Kaplan RS, Matsumura EM. Management accounting. Pearson Prentice Hall: Upper Saddle River; 2007.
12. Mintzberg H. The structuring of organizations. Prentice-Hall: Englewood Cliffs; 1979.
13. Conferenza delle Regioni e delle Provincie Autonome. Conferenza delle Regioni e delle Provincie Autonome: Roma; 2021.
14. Ministero della Salute. Decreto 18 ottobre 2012. Remunerazione prestazioni di assistenza ospedaliera per acuti, assistenza ospedaliera di riabilitazione e di lungodegenza post acuzie e di assistenza specialistica ambulatoriale. (13A00528). Gazzetta Ufficiale della Repubblica Italiana [Internet]. Serie Generale, n. 23 del 28 gennaio 2013b. Allegato 3 Prestazioni di assistenza specialistica ambulatoriale.Publisher Full Text
15. Brouwer W, Rutten F, Koopmanschap M. Economic evaluation in health care: merging theory with practice. Oxford University Press: Oxford; 2001.
16. Paolini D, Tiseo M, Demma F. Ventana ALK (D5F3) in the detection of patients affected by anaplastic lymphoma kinase-positive non-small-cell lung cancer: Clinical and Budget Effect. Clin Lung Cancer. 2018; 19:e735-e743. DOI
17. Pisapia P, Pepe F, Baggi A. Next generation diagnostic algorithm in non-small cell lung cancer predictive molecular pathology: The KWAY Italian multicenter cost evaluation study. Crit Rev Oncol Hematol. 2022; 169:103525. DOI