Budget impact analysis of breast cancer screening in Italy: The role of new technologies

Abstract

Although mammography screening significantly reduces breast cancer mortality, women could present different morphological characteristics that do not allow the correct vision of their breasts and the detection of cancer, resulting in a delay of diagnosis and an increase in the risk of mortality. The present study aims at analyzing potential areas of improvement of the current screening programs and then hypothesizing alternative technologies to use within the diagnostic phase, from an economic point of view. A Budget Impact Analysis approach was implemented, considering the Italian National Healthcare Service perspective, and representing the healthcare expenditure evolution, over three years. In the Budget Impact Analysis model, two distinct phases of the screening programs were considered: (1) the screening/diagnosis phase and (2) the phase related to cancer care and treatments of patients. The results provide clinicians and policy makers with a rational method to forecast economic resources in the screening programs in a general context of limited resources. In particular, results of the Budget Impact Analysis showed that, while the introduction of the ABUS InveniaTM technology into the screening programs would lead to an increase in the screening phase expenditure, it would generate an economic advantage related to the patients treatment and care.

Keywords

breast cancer budget impact analysis Italy screening programs

Introduction

Mammography is the technology mostly used in the breast cancer screening programs, proposed to the general population, and leading to a significant reduction in mortality, in the specific target group of women attending the program.¹

However, the sensitivity of mammography is highly variable, depending on the breast density. In women presenting dense breasts, Kolb et al.² report that mammographic sensitivity decreases from 85 to 48%. Furthermore, in this type of breast, a negative mammogram report does not exclude the presence of a neoplasm.

Breast density not only represents an independent risk factor for developing a cancer (four to six times increased risk, over lifetime), but also a serious limitation for the interpretation of the screening test.^3,4 The final result could be a diagnostic delay, but also an increase in the mortality risk.⁵

In this view, an efficient diagnostic process and an early diagnosis could be crucial factors, both from a clinical perspective and from an economic point of view.

To overcome the above-mentioned limitations, other imaging techniques should be used as additional screening tests for women presenting dense breast, in particular with regard to the availability of ultrasound technique in the clinical practice.

Ultrasound is a non-ionizing radiation imaging technique that is usually used in the surveillance of both high-risk⁶ and medium-risk women presenting dense breasts.^7,8 The ultrasound technique is able to detect invasive cancers, not visible at mammography, often of small size, and with a favorable prognosis.

Ultrasound obtains a diagnostic increase by 50% (detection rate 8, per 1000 patients) if compared with mammography.⁶ However, it is an operator-dependent imaging technique, and it requires an important execution time. The ACRIN6666 study⁶ reports around 20 min, on average, to execute an adequate ultrasound exam, procedure that requires an intense professional training, and also it needs a steep learning curve. Ultrasound implies an increase in the recall and false-positive rate. These factors lead to a significant rise in costs, duplication of exams, and, consequently, organizational issues.

Even though the mammography screening significantly reduces the breast cancer mortality,⁹ in general, women, presenting different morphological characteristics, may not allow the correct vision of their breasts and the detection of cancers, resulting in the delay of diagnosis, increasing the risk of mortality.⁵

As a result, in the present study, a different screening pathway was approached and proposed, consisting of a specific device for the automatic acquisition of ultrasound images (Automated Breast UltraSound, ABUS Invenia™).^10,11 The examinations could be performed by a trained radiology technician, thus releasing the radiologist from problems related to timing (in this case the interpretation and reporting could be postponed at another time) and reproducibility (due to the automatic images acquisition).

ABUS Invenia™ was approved by Food and Drug Administration, as a screening system for women with dense breast, after the good results reported in the prospective multicenter observational study SomoInsight (an increase in the detection of 1.9 per 1000 women screened, with an increase in sensitivity equal to 26.7%).¹² Furthermore, 93% of the tumors were invasive, node-negative, and less than 2 cm, in 87% of cases.¹²

In this view, the introduction of new technologies in the screening programs should require an in-depth analysis, useful to define their economic and organizational sustainability, in general settings characterized by a paucity of resources.

This is particularly relevant, since screening programs pathways often depend on the Hospitals internal organizations, their resources availability, or the peculiar characteristics of the organizations involved in the screening activities (for example, the presence of different type of screening technologies).

Thus, the implementation of lean screening programs, requiring lower resources and guaranteeing a greater effectiveness, could be a feasible solution for all Healthcare Systems, both the National Healthcare Service (NHS) and private insurance-based services.

Moving on from these premises, the present study aimed at analyzing the potential areas of improvement and optimization of the current screening programs, hypothesizing the panel of technologies to be used within the diagnostic phase.

Materials and methods

To achieve the above-mentioned objective, a Budget Impact Analysis (BIA) approach was developed in order to estimate and predict the economic and financial consequences referring to the adoption and diffusion of new technologies into a healthcare system with finite resources,^13,14 after the collection of data related to the real-world development of the patients screening clinical pathway.

The study design was structured in three different steps as follows:

Literature review and data retrieval from institutional registers, at National and Regional level, considering the Italian NHS. In particular, this first step was useful to gather data inputs for the BIA model. The cancer detection rate was extracted from literature evidence, while the population attending the screening programs was derived from National registers. The prevalence and incidence rates of the pathology, the annual invites to screening programs, as well as the rate of breast cancer in women who conducted the screening, were retrieved from Regional registers. Specifically, authors considered the Lombardy Region registers because these kinds of information were not available and, especially, updated at National level. Regional data were used in the BIA model, even though the Italian NHS perspective was assumed. However, Lombardy Region data are representative for the entire National context, being the first Italian Region, for healthcare sector expenditure (about 17 million euro), and for inhabitants (about 10 million).

Collection of real-world data. This step was useful to gather data inputs for the BIA model, with regard to specific data not available from National and Regional registries. An observational study was conducted in two prestigious hospitals (Gavazzeni Hospital in Bergamo and Fondazione IRCCS Istituto Nazionale dei Tumori in Milan) for the care of breast cancer due to the presence of a specific ward for this pathology called Breast Unit. Retrieved data were related to the percentage of the use of technologies in the screening programs and the incidence of the T1 stage and T2–T3–T4 stages of breast cancer in the population taken in charge. Moreover, a sample of 100 women affected by cancer was investigated in order to map the real clinical pathways and the resource absorption.

Design of Budget Impact Analysis model. The Italian NHS perspective was taken into account, and the healthcare expenditure evolution, up to three years, was represented. Two distinct phases of the screening program were considered: (i) the screening/diagnosis phase and (ii) the phase related to the cancer care and patients’ treatment. The following three different scenarios, related to the technologies potentially used in the screening pathways, were analyzed.

AS–IS scenario. Territorial Hospitals or Services contacted and invited to participate in screening programs target female population. A mammography is performed, and, if the result is positive, women will be contacted for further examinations (i.e. ultrasound).

IDEAL scenario. This scenario is similar as previously described, but an additional ultrasound approach was guaranteed to all the women participating in the screening program and having specific characteristics (the dense breast, with BI RADS 3 and 4). This scenario represents the best solution that could be adopted in the field of screening programs to optimize the clinical pathway. However, it is not currently considered and applied because of its economic and organizational unsustainability.

TO BE scenario. This scenario considers the introduction of the ABUS Invenia™ technology in addition to mammography and ultrasound. It should be noted that both ABUS Invenia™ and ultrasound were performed in a specific female target population: characterized by dense breast (BI RADS 3 and 4) and an incremental cancer detection rate.

The BIA model inputs are described in the following:

Population attending the screening programs: Data related to the population entering the first year of the model derived from National database (ISTAT¹⁵). Women aged between 50 and 70 years in Italy amounted to 8,529,765 subjects (year of reference: 2018), thus representing the population invited to attend the screening programs within 24 months. This data, for all considered scenarios, was adjusted with the percentage of annual invites equal to 46.58%. Actually, only 46.42% of women carried out a mammographic exam. These percentages of impact were derived from the Observatory on Oncological Screening in ombardy Region data.¹⁶ The total number of annual invites and the total number of women carrying out mammographic exams were considered and calculated as percentage on the overall regional target population. In the second and third year of analysis, the population entering the model was calculated, with the same method, considering the national estimated annual growth rate (2.09%) and mortality index (0.02%)¹⁵ in order to define the correct number of women aged between 50 and 70 years. In this view, the population entering the model was equal to 1,850,925; 1,870,267; and 1,889,811 women, respectively, in the first, second, and third year of analysis.

2. Recall rate: In all the considered scenarios, 100% of women carried out a mammographic exam, while the rate of recalls for the ultrasound differed. In particular, the AS–IS scenario supposed that the rate of recalls for the ultrasound for women who presented a dense breast was equal to 7.23% based on observational data collected in the two Hospitals involved in the analysis. The IDEAL scenario hypothesized that the rate of recalls for the ultrasound was equal to 49.54%. This percentage considers both the previously mentioned rate of recall (AS-IS scenario, 7.23) and the incidence of women who presented a dense breast with BI RADS categories 3 and 4 (equal to 42.31%). In the TO BE scenario (i.e. the introduction of the ABUS Invenia™ technology), the recall rate with an ultrasound diagnostic examination was equal to 7.42% (deriving from the 7.23%, related to the AS-IS scenario, increased by 0.19%, represented by the incremental cancer detection rate of 1.9 cancers for 1000 screened women, as reported in the literature evidence¹²). This meant that the ABUS Invenia™ technology was able to identify more cancers (+0.19%), if compared with the scenario characterized by the only use of mammography. Furthermore, the TO BE scenario hypothesized to use the ABUS Invenia™ technology on 42.31% of screened women with mammography due to the presence of a dense breast (BI RADS categories 3 and 4).

3. Population affected by breast cancer: After the screening program, the diagnosed female population, with a positive exam, detecting the presence of a cancer continues the clinical pathway with the phase of cancer care and patients’ treatment. For each year analyzed, the entered female population derived from the screened one, as above described. The women affected by breast cancer amounted to 7.73% of screened female population (Screening Registry of the Lombardy Region, 2015). The population affected by breast cancer, in terms of number of target female patients entering the model, was the same for all the three analyzed scenarios. The distribution of the eligible population in the different stages (from T1 to T4) of cancer pathology was different, depending on the scenario under analysis. In the AS IS scenario, 50% of women were ascribed in T1 stage and the remaining 41% in T2–T3–T4 stages, according to real-world data. In the IDEAL scenario, literature evidence was considered.¹⁷ The mammography associated with an ultrasound recall may increase the probability of identification of cancers at an earlier stage to 15.9%. In this view, the number of women at T1 stage was greater in IDEAL scenario, if compared with the AS–IS scenario. In the TO BE scenario, the ABUS Invenia™ technology¹² may increase the probability of cancers identification at an earlier stage to 55%. The number of women at T1 stage was greater in TO BE scenario than in the AS–IS and IDEAL scenarios.

3. Economic evaluation: An average management cost of cancer care and patients’ treatment expenditure was derived from the data collected in the observational analysis. In particular, the total amount of hematologic exams, diagnostic and surgical procedures, outpatients and medical examinations, and hospital admissions were investigated for all the possible cancer stages. To define the different scenarios, pathways, and the granulation of needs, the process mapping technique was initially applied,¹⁸ and, subsequently, the Activity-Based Costing-ABC technique¹⁸ was performed, aimed at measuring the costs and performance of activities. In particular, information was evaluated in accordance with the Lombardy Region’s outpatients and hospital admissions Reimbursement Tariffs and the NHS official drugs price list. The economic inputs, distinguishing the population at T1 stage and at T2–T3–T4 stages, are reported in Table 1.

4. Scenario and sensitivity analysis: Scenarios analyses were performed, changing the parameters who affected population entering the model, as described in the followings.

The percentage of annual invites equal to 57% according to national average invite rate¹⁹ was used, changing the data retrieved by the Lombardy Region register.¹⁵ This scenario modified the population attending the screening programs: in this case, the female population entering the model in the first year of analysis was equal to 2,771,321 women.

The percentage of women presenting a dense breast (BI RADS categories 3 and 4) equal to 25% was used, according to the results of the observational study conducted in the hospitals involved in the analysis.

The number of women affected by breast cancer, detected during the screening phase, was modified, assuming an increase of +5% of the previous value (7.73%¹⁶).

The distribution of the eligible population in the different stages (from T1 to T4) was modified in the IDEAL Scenario. The recall rate of the mammography associated with ultrasound, at an earlier stage, was moved from 15.9¹⁷ to 28.5%.² The scenario and sensitivity analyses were performed to verify the robustness of the results. Bayesian statistics were implemented, since Bayesian methods provide a complete paradigm for both statistical inference and decision making under uncertainty. Gamma distributions were accordingly developed in order to verify the robustness and variability of the results in the presence of uncertainty factors (clinical pathway costs). In particular, gamma distributions considered, as parameters, the mean value and the standard deviation for each scenario under assessment. The probability to have any level of average cost value, represented by the gamma distributions of the TO BE scenario, lower with respect to both the AS IS scenario and the IDEAL scenario, was analyzed, defining the sustainability of the models proposed and the optimization of the pathways under evaluation.

Results

The number of services performed (in terms of mammogram, ultrasound, and ABUS Invenia™ technologies) to the eligible female population, considered into the Budget Impact Analysis for the screening phase, was reported (see Table 2).

Table 1.

Economic evaluation of the care pathway of women affected by breast cancer, in the first, second, and third years of analysis, T1 and T2–T3–T4 stages.

	Patient at T1 stage	%	Patient at T2–T3–T4 stages	%
First year after breast cancer diagnosis
Therapy cost	€5413.87	57.84	€5679.35	51.60
Laboratory tests cost	€1042.63	11.4	€1900.69	17.27
Diagnostic and specialist procedures cost	€186.56	1.99	€199.38	1.81
Hospitalization	€2716.82	29.03	€3227.63	29.32
Total first year	€9359.88	100.00	€11,007.05	100.00
Second and third year after breast cancer diagnosis
Therapy cost	€473.43	37.50	€473.43	37.50
Laboratory tests cost	€610.06	48.33	€610.06	48.33
Diagnostic and specialist procedures cost	€178.87	14.17	€178.87	14.17
Total second and third year	€1,262.36	100.00	€1,262.36	100.00

Data source: sample of 100 patients in the hospital setting of the Fondazione IRCCS Istituto Nazionale dei Tumori in Milan.

Table 2.

Number of services (mammogram, ultrasound, ABUS Invenia™ technology) performed during the he screening/diagnosis phase over the three-year time horizon in the three scenarios (AS-IS, IDEAL and TO BE).

	AS IS scenario	IDEAL scenario	TO BE scenario
First year
Population: 1,850,925 women who conducted the screening
Mammogram (services/year)	1,850,925	1,850,925	1,850,925
Ultrasound (services/year)	133,822	916,948	137,339
ABUS (services/year)	–	–	783,126
Second year
Population: 1,870,267 women who conducted the screening
Mammogram (services/year)	1,870,267	1,870,267	1,870,267
Ultrasound (services/year)	135,220	926,530	138,774
ABUS (services/year)	–	–	791,310
Third year
Population: 1,889,811 women who conducted the screening
Mammogram (services/year)	1,889,811	1,889,811	1,889,811
Ultrasound (services/year)	136,633	936,213	140,224
ABUS (services/year)	–	–	799,579

On the basis of the services performed, the economic impact was evaluated, applying the following tariffs for each procedure carried out: €44.87 for mammogram, €46.80 for ultrasound, and €44.87 for ABUS Invenia™ technology. Since the Italian NHS has not yet determined the reimbursement tariff for the examinations performed with ABUS Invenia™ technology, the hypothesis implemented in the model was that the ABUS examinations were refunded with the same tariff used in the case of mammography (see Table 3).

Table 3.

Economic impact (mammogram, ultrasound, ABUS InveniaTM technology) of the screening/diagnosis phase over the three-year time horizon in the three scenarios (AS-IS, IDEAL, and TO BE).

	AS IS scenario	IDEAL scenario	TO BE scenario	% IDEAL-AS IS	% TO BE-AS IS
First year
Mammogram	€83,051,000.90	€83,051,000.90	€83,051,000.90	0.00%	0.00%
Ultrasound	€6,262,863.58	€42,913,175.87	€6,427,447.82	585.20%	2.63%
ABUS	–	–	€35,138,878.48	–	–
Total economic impact	€89,313,864.47	€125,964,176.77	€124,617,327.19	41.04%	39.53%
Second year
Mammogram	€83,918,883.85	€83,918,883.85	€83,918,883.85	0.00%	0.00%
Ultrasound	€6,328,310.50	€43,361,618.56	€6,494,614.65	585.20%	2.63%
ABUS	–	–	€35,506,079.76	–	–
Total economic impact	€90,247,194.36	€127,280,502.42	€125,919,578.26	41.04%	39.53%
Third year
Mammogram	€84,795,836.19	€84,795,836.19	€84,795,836.19	0.00%	0.00%
Ultrasound	€6,394,441.35	€43,814,747.48	€6,562,483.37	585.20%	2.63%
ABUS	–	–	€35,877,118.29	–	–
Total economic impact	€91,190,277.54	€128,610,583.67	€127,235,437.86	41.04%	39.53%
Total economic impact in the considered time horizon	€270,751,336.36	€381,855,262.85	€377,772,343.31	41.04%	39.53%

Over the three years of analysis, the IDEAL scenario absorbed more NHS resources (+41% if compared to the AS–IS scenario), considering the only screening phase. The TO BE scenario absorbed an amount equal to + 39.5%, if compared with the AS–IS scenario. An increasing trend of resources absorption was defined, from the first to the third year, due to demographic dynamics (i.e. the increase in the population to be potentially invited to participate in screening programs).

Considering the phase related to cancer care and patients’ treatment, the economic values, reported in Table 1, were multiplied for the number of women, according to the different detection cancer stages, as previously described, and to the year of reference. Table 4 shows that, over the three years of analysis, the more expensive scenario is the AS–IS, while the IDEAL and the TO BE scenario allowed an economic saving (respectively, –0.95% and –3.29%), due to the early detection of the breast cancer.

Table 4.

Economic impact of the phase of care pathway of women affected by cancer over the three-year time horizon in the three scenarios (AS-IS, IDEAL, and TO BE).

	AS IS scenario	IDEAL scenario	TO BE scenario	% IDEAL-AS IS	% TO BE-AS IS
First year
Population: 1,850,925 women who conducted the screening
	€1,435,810,631.33	€1,420,447,330.37	€1,382,667,137.47	–1.07%	–3.70%
Second year
Population: 144,572 women who are diagnosed with breast cancer in the second year + 143.077 women affected by breast cancer (diagnosis in the first year)
	€1,631,427,691.19	€1.615.903.860,35	€1,577,728,905.27	–0.95%	–3.29%
Third year
Population: 146,083 women who are diagnosed with breast cancer in the third year + 143.077 women affected by breast cancer (diagnosis in the first year) + 144.572 women affected by breast cancer (diagnosis in the second year)
	€1,829,092,539.90	€1,813,406,461.12	€1,774,832,518.98	–0.86%	–2.97%
Total economic impact in the considered time horizon	€4,896,330,862.41	€4,849,757,651.85	€4,735,228,561.72	–0.95%	–3.29%

The final results of the budget impact analyses performed in the three scenarios considered, were obtained, as the sum of the screening phase costs and the cancer care and patients’ treatment phase expenditure, divided into the three-year time horizon (see Table 5).

Table 5.

Economic impact of the two considered phases (screening and care pathway of women affected by cancer), divided into the three-year time horizon in the three scenarios (AS-IS, IDEAL and TO BE).

	AS IS scenario	IDEAL scenario	TO BE scenario	% IDEAL-AS IS	% TO BE-AS IS
First year	€1,525,124,495.80	€1,546,411,507.14	€1,507,284,464.67	1.40%	–1.17%
Second year	€1,721,674,885.55	€1,743,184,362.77	€1,703,648,483.53	1.25%	–1.05%
Third year	€1,920,282,817.43	€1,942,017,044.79	€1,902,067,956.84	1.13%	–0.95%
Total economic impact	€5,167,082,198.78	€5,231,612,914.70	€5,113,000,905.03	1.25%	–1.05%

Results suggested that the IDEAL scenario need incremental investments (+1.25%, if compared with the AS–IS scenario), while the TO BE scenario was the more acceptable for the Healthcare Service (–1.05% compared to the AS-IS scenario), thus allowing a potential economic saving equal to more than €54 million for the Italian NHS.

All the scenario analyses (Table 6) confirmed that the TO BE scenario allowed economic savings (ranging from –1.89% to –1.05% respect the AS IS scenario), while the IDEAL scenario absorbed more resources (ranging from +0.53% to +1.25% respect the AS IS scenario).

Table 6.

Scenario analysis: Scenario A considered percentage of annual invites equal to 57%; Scenario B considered percentage of women presenting a dense breast (BI RADS categories 3 and 4) equal to 25%; Scenario C considered a percentage of the women affected by breast cancer equal to 8.12%, after the screening phase; Scenario D considered the probability of 28.5% of identification of cancers at an earlier stage in the IDEAL Scenario.

	AS IS scenario	IDEAL scenario	TO BE scenario	% IDEAL-AS IS	% TO BE-AS IS
Scenario A
First year	€2,283,513,463.26	€2,315,385,692.13	€2,256,802,192.80	1.40%	–1.17%
Second year	€2,577,797,121.31	€2,610,002,483.78	€2,550,806,956.08	1.25%	–1.05%
Third year	€2,875,161,385.25	€2,907,703,296.76	€2,847,889,183.17	1.13%	–0.95%
Total economic impact	€7,736,471,969.82	€7,833,091,472.67	€7,655,498,332.05	1.25%	–1.05%
Scenario B
First year	€1,525,124,495.80	€1,531,417,016.34	€1,492,908,336.41	0.41%	–2.11%
Second year	€1,721,674,885.55	€1,728,033,179.54	€1,689,122,124.74	0.37%	–1.89%
Third year	€1,920,282,817.43	€1,926,707,531.69	€1,887,389,797.59	0.33%	–1.71%
Total economic impact	€5,167,082,198.78	€5,186,157,727.57	€5,069,420,258.74	0.37%	–1.89%
Scenario C
First year	€1,596,916,532.65	€1,617,435,362.84	€1,576,419,271.11	1.28%	–1.28%
Second year	€1,803,250,473.55	€1,823,983,716.28	€1,782,538,983.67	1.15%	–1.15%
Third year	€2,011,736,633.44	€2,032,686,572.97	€1,990,808,827.51	1.04%	–1.04%
Total economic impact	€5,411,903,639.64	€5,474,105,652.09	€5,349,767,082.29	1.15%	–1.15%
Scenario D
First year	€1,525,124,495.80	€1,534,236,815.83	€1,507,284,464.67	0.60%	–1.17%
Second year	€1,721,674,885.55	€1,730,882,459.08	€1,703,648,483.53	0.53%	–1.05%
Third year	€1,920,282,817.43	€1,929,586,567.27	€1,902,067,956.84	0.48%	–0.95%
Total economic impact	€5,167,082,198.78	€5,194,705,842.18	€5,113,000,905.03	0.53%	–1.05%

The sensitivity analysis also confirmed that the results could be considered as robust (Figure 1). Figure 1 reports the comparison between the considered scenarios (AS IS, IDEAL e TO BE) in order to define in how many observations (to be considered on the X axis), the results deriving from the TO BE scenario, represented the best solution and the most convenient economic landscape for the Healthcare System. In particular, any time the green curve concerning the TO BE scenario is represented under the other curves, the TO BE scenario is the preferable one and the most sustainable. Specifically, comparing the IDEAL and the AS IS scenario, the IDEAL scenario reported a probability of 18% to absorb fewer resources than the AS IS scenario. While comparing the TO BE scenario and the AS IS scenario, the TO BE scenario has a probability of 77% to absorb fewer resources than the AS IS one. In conclusion, comparing the TO BE scenario and the IDEAL scenario, the TO BE scenario has a probability of 80% to absorb fewer resources than the IDEAL scenario.

Figure 1.

Sensitivity analysis. Gamma distributions were developed in order to verify the robustness of the results in the presence of uncertainty factors. The X axis reported the number of observations, while the Y axis reported the total economic impact (the sum of three years of BIA) of the two considered phases (screening and care pathway of women affected by cancer).

Discussion

As life expectancy continues to rise, the cancer burden is also predicted to increase, thus representing a relevant economic concern.

Only strategies related to the prevention, the screening, and the early detection of new cancers, in the female population, could help to contain or lower this significant economic and clinical burden.²⁰

In this view, literature evidence is poor, in relation with the economic impact of technologies used in breast screening programs and investment planning related to efficient healthcare services resources utilization for the diagnostic setting. Thus, it becomes an urgent priority to give practical information to scholars and practitioners with regard to the organizational and economic definition of proper screening programs.

This aspect already plays a key role in Hungary,²¹ Russia,²² New Zealand,²³ Costa Rica and Mexico,²⁴ Holland,²⁵ Canada,²⁶ etc., in particular, because of budget constraints, greater healthcare production, and the introduction of national screening programs.

The results of the present study provide clinicians and policy makers with a rational method to forecast economic resources regarding screening programs in a general context of limited resources. Indeed, information, in terms of budget impact analysis, should help decision-makers to fully understand the financial impact of such screening programs on Italian NHS expenditure.

Literature reported different methods to evaluate and analyze the problem of limited resources management in the setting of breast cancer. Cost-effectiveness analysis seems to be the most suggested approach that should be taken into account^21–24 as it allows for the identification of the most efficient mix of interventions to improve the general population’s health.

In addition, simulation and budget impact analysis are reported in literature of breast cancer screening, achieving different aims. Koleva-Kolarova et al.²⁵ evaluated the quantitative benefits, harms, and cost-effectiveness of lowering the starting age of breast cancer screening in the Dutch general population. Carter et al.²⁷ used a Monte Carlo computer simulation to study the balance of benefits and harms of mammographic breast cancer screening for average-risk women, comparing two different screening strategies (mixed annual-biennial screening guidelines vs. biennial guidelines). A discrete-event simulation model was developed by Comas et al.²⁸ to assess the budgetary impact of switching from screen-film mammography to full-field digital mammography, in the Spanish population.

According to the literature evidence mentioned above, the present study contributes to the current debate on breast screening programs from an economic perspective. The results focused their attention on the role of the different mix of existing technologies or the introduction of technology innovation, able to early diagnose women affected by cancer, exceeding the traditional cost-effectiveness approach used to compare different technologies, and thus allow the selection of the most promising one.

The results assume the importance of sustainability aspects, providing additional data supporting the need to redesign the actual pathways proposed in current Italian screening programs, at least for some groups of women. Moreover, in the present study, the analysis was not limited to the diagnostic phase (as in Arrospide et al.²⁹), but, it is presented as a holistic approach, considering also the phase of care and treatment of breast cancer patients.

Another important contribution is related to the possibility to create awareness of the need to improve the current screening programs, hypothesizing different technologies to use within the diagnostic phase, estimating the feasibility of this change in a medium time horizon (i.e. 3 years), and not simulating a wide time horizon (i.e. 20 years), as happened applying Markov models projections.

In particular, the ABUS Invenia™ technology could be introduced as a support to mammogram examination, especially for women presenting dense breasts, though not totally replacing it, thus proposing a new approach in screening programs.

The BIA model showed that the introduction of the ABUS Invenia™ technology in screening programs could lead to an increase in the screening phase expenditure, but could stimulate a significant decrease in the female patients’ care and cancer treatment phase. The ability of the technology in identifying a greater number of cancers, especially at an early stage, has a positive impact on the mortality rate of pathology, thus not only empowering diagnostic phase, but also optimizing the outcome results and giving a better final care to patients.

Considering both the screening phase and the women affected by cancer care and treatment phase, the TO BE scenario, related to the introduction of the ABUS Invenia™ technology, suggested a lower NHS resources absorption.

Focusing on the managerial implications, the present study provides practitioners with a deeper understanding concerning factors influencing screening programs, useful to improve performance and optimization of resources pathway. The results could support several categories of stakeholders, such as hospital managers, professionals responsible for resources allocation (e.g. staff, investments, etc…), and regional or national healthcare system managers to improve the overall policy for breast cancer screening programs.

However, despite the relevance of the results, the model has some limitations.

Data related to ultrasound recalls rate and process mapping of treatment and care pathway of women affected by cancer were retrieved from real-world data, based on observational studies conducted within two hospitals, not yet published though more accurate and valid for the Italian and European setting of reference. Another limitation is the lack of recent information, within the national setting, related to both the rate of annual invites to participate in screening programs and information based on evidence provided by the Lombardy Region.

Therefore, it should be a topic for future research to update the results, considering a multi-dimensional approach, such as the one proposed by Health Technology Assessment methodology.³⁰

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Pisano

Gatsonis

Hendrick

, et al. Diagnostic performance of digital versus film mammography for breast-cancer screening, October 27, 2005. N Engl J Med 2005; 353: 1773–1783.

Kolb

Lichy

Newhouse

JH.

Comparison of the performance of screening mammography, physical examination, and breast US and evaluation of factors that influence them: an analysis of 27,825 patient evaluations. Radiology 2002; 225: 165–175.

Boyd

Guo

Martin

, et al. Mammographic density and the risk and detection of breast cancer. N Engl J Med 2007; 356: 227–236.

Freer

PE.

Mammographic breast density: impact on breast cancer risk and implications for screening. Radiographics 2015; 35: 302–315.

Rivera-Franco

Leon-Rodriguez

Delays in breast cancer detection and treatment in developing countries.

Breast Cancer (Auckl). 8 January 2018. DOI: 10.1177/1178223417752677.

Berg ACRIN 6666 Investigators. Combined screening with ultrasound and mammography vs mammography alone in women at elevated risk of breast cancer. JAMA 2008; 299: 2151–2163.

Brem

Lenihan

Lieberman

, et al. Screening breast ultrasound: past, present, and future. AJR Am J Roentgenol 2015; 204: 234–240.

Hooley

Scoutt

Philpotts

LE.

Breast ultrasonography: state of the art. Radiology 2013; 268: 642–659.

Kerlikowske

Grady

Rubin

, et al. Efficacy of screening mammography: a meta-analysis. JAMA 1995; 273: 149–154.

10.

Kaplan

SS.

Automated whole breast ultrasound. Radiol Clin N Am 2014; 52: 539–546.

11.

Kelly

Richwald

GA.

Automated whole-breast ultrasound: advancing the performance of breast cancer screening. Semin Ultrasound CT MR 2011; 32: 273–280.

12.

Brem

Tabár

Duffy

, et al. The SomoInsight study assessing improvement in detection of breast cancer with three-dimensional automated breast US in women with dense breast tissue: the SomoInsight Study. Radiology 2015; 274: 663–673.

13.

Mauskopf

, et al. Principles of good practice for budget impact analysis: report of the ISPOR task force on good research practices-budget impact analysis. Value Health 2007; 10(5): 336–347.

14.

Sullivan

Mauskopf

Augustovski

, et al. Principles of good practice for budget impact analysis II: report of the ISPOR task force on good research practices – budget impact analysis. Value Health 2014; 17: 5–14.

15.

ISTAT. Italian Institute of Statistics, www.istat.it/ (2018, accessed 13 April 2019).

16.

Cereda

Observatory on oncological screening in Lombardy Region (based on activities performed in year of reference: 2014), www.regione.lombardia.it/wps/wcm/connect/a2dbbbb5-9e9b-4e1a-b54b-6899b0787a 76/Report_screening_2015.pdf?MOD=AJPERES&CAC HEID=a2dbbbb5-9e9b-4e1a-b54b-6899b0787a76 (2016, accessed 10 January, 2019).

17.

Schaefer

Waldmann

Katalinic

, et al. Influence of additional breast ultrasound on cancer detection in a cohort study for quality assurance in breast diagnosis: analysis of 102,577 diagnostic procedures. Eur Radiol 2010; 20: 1085–1092.

18.

Vagnoni

Potenta

L'activity based costing in sanità: il caso dell'ossigenoterapia. Mecosan 2003; 47: 149–161.

19.

Observatory on National Screening Program (based on activities performed in year of reference: 2013), www.osservatorionazionalescreening.it/sites/default/files/allegati/Screening_2014_web.pdf (2014, accessed 08 April, 2019).

20.

Knox

Health economic decision making in Europe – a new priority for breast cancer advocacy. Breast 2009; 18: 71–72.

21.

Noémi

Tamás

Piroska

, et al. Economic evaluation of population-based mass screening for the early detection of cancer: a systematic review. Magyar Onkologia 2002; 46(2): 119–129.

22.

Orlov

Alferenko

Zelënyǐ

AV.

Economic aspects of specialized preventive screening for breast cancer. Voprosy Onkol 2002; 48(2): 244–246.

23.

Menon

Devlin

Richardson

AK.

The costs of mammography screening in New Zealand: evidence from the pilot programmes, New Zealand. Med J 1994; 107(991): 501–503.

24.

Niëns

Zelle

Gutiérrez-Delgado

, et al. Cost-effectiveness of breast cancer control strategies in Central America: the cases of Costa Rica and Mexico. PlosOne 25 April 2014. DOI: 10.1371/journal.pone.0095836.

25.

Koleva-Kolarova

Daszczuck

de Jonge

, et al. A modelling study to evaluate the costs and effects of lowering the starting age of population breast cancer screening. Maturitas 2018; 109: 81–88.

26.

Nikitovic-Jokic

Palimaka

, et al. Ultrasound as an adjunct to mammography for breast cancer screening. Ont Health Technol Assess Ser 2016; 16(15): 1–71.

27.

Carter

Castro

Morcos

RN.

Insights into breast cancer screening: a computer simulation of two contemporary screening strategie. AJR 2018; 210(3): 564–571.

28.

Comas

Arrospide

Mar

, et al. Budget impact analysis of switching to digital mammography in a population-based breast cancer screening program: a discrete event simulation model. PlosOne 2014; 9(5): e97459.

29.

Arrospide

Rue

van Ravesteyn

, et al. Economic evaluation of the breast cancer screening programme in the Basque Country: retrospective cost-effectiveness and budget impact analysis. BMC Cancer. Epub ahead of print 1 June 2016. DOI: 10.1186/s12885-016-2386-y.

30.

Drummond

Schwartz

Jönsson

, et al. Key principles for the improved conduct of health technology assessments for resource allocation decisions. Int J Technol Assess Health Care 2008; 24: 244–258.