Journal of Hospital Infection 143 (2024) 8e17

Available online at www.sciencedirect.com

Journal of Hospital Infection

journal homepage: www.elsevier.com/locate/jhin

Estimating the savings of a national project to prevent

healthcare-associated infections in intensive care units
R.M.C. Oliveira c, A.H.F. de Sousa c, M.A. de Salvo c, A.J. Petenate f,
A.K.F. Gushken b, E. Ribas d, E.M.S. Torelly e, K.C.C.D. Silva a, L.M. Bass c,
P. Tuma c, f, P. Borem f, L.Y. Ue g, C.G. de Barros c, S. Vernal a, b, c, d, e, *, on behalf
of the Saúde em Nossas Mãos Collaborative Study Group
a
Hospital Alemão Oswaldo Cruz, Sao Paulo, Brazil
b
Hcor, Sao Paulo, Brazil
c
Hospital Israelita Albert Einstein, Sao Paulo, Brazil
d
Hospital Moinhos de Vento, Porto Alegre, Brazil
e
Hospital Sı´rio-Libanês, Sao Paulo, Brazil
f
Institute for Healthcare Improvement, Cambridge, MA, USA
g
Ministério da Saúde, Brasilia, Brazil

A R T I C L E I N F O S U M M A R Y

Article history: Background: Healthcare-associated infections (HAIs) have a significant impact on

Received 23 June 2023 patients’ morbidity and mortality, and have a detrimental financial impact on the
Accepted 3 October 2023 healthcare system. Various strategies exist to prevent HAIs, but economic evaluations are
Available online 6 October 2023 needed to determine which are most appropriate.
Aim: To present the financial impact of a nationwide project on HAI prevention in
Keywords: intensive care units (ICUs) using a quality improvement (QI) approach.
Bloodstream infection Methods: A health economic evaluation assessed the financial results of the QI initiative
Healthcare-associated ‘Saúde em Nossas Mãos’ (SNM), implemented in Brazil between January 2018 and December
infections 2020. Among 116 participating institutions, 13 (11.2%) fully reported the aggregate cost and
Intensive care unit stratified patients (with vs without HAIs) in the pre-intervention and post-intervention
Financial management periods. Average cost (AC) was calculated for each analysed HAI: central-line-associated
Quality improvement bloodstream infections (CLABSIs), ventilator-associated pneumonia (VAP) and catheter-
Prevention and control associated urinary tract infections (CAUTIs). The absorption model and time-driven activ-
Urinary tract infection ity-based costing were used for cost estimations. The numbers of infections that the project
Ventilator-associated could have prevented during its implementation were estimated to demonstrate the
pneumonia financial impact of the SNM initiative.
Results: The aggregated ACs calculated for each HAI from these 13 ICUs e US$8480 for
CLABSIs, US$10,039 for VAP, and US$7464 for CAUTIs e were extrapolated to the total
number of HAIs prevented by the project (1727 CLABSIs, 3797 VAP and 2150 CAUTIs). The
overall savings of the SNM as of December 2020 were estimated at US$68.8 million, with an
estimated return on investment (ROI) of 765%.
Conclusion: Reporting accurate financial data on HAI prevention strategies is still chal-
lenging in Brazil. These results suggest that a national QI initiative to prevent HAIs in

* Corresponding author. Address: Escritório de Excelência Einstein. Av. Paulista, 2300 e Sao Paulo/SP, 01310-200, Brazil.
E-mail address: vernal.carranza@gmail.com (S. Vernal).

https://doi.org/10.1016/j.jhin.2023.10.001
0195-6701/ª 2023 The Authors. Published by Elsevier Ltd on behalf of The Healthcare Infection Society. This is an open access article
under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 R.M.C. Oliveira et al. / Journal of Hospital Infection 143 (2024) 8e17 9
critical care settings is a feasible and value-based approach, reducing financial waste and
yielding a significant ROI for the healthcare system.
ª 2023 The Authors. Published by Elsevier Ltd
on behalf of The Healthcare Infection Society. This is an open access article
under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Background Methods
The World Health Organization has recognized healthcare- Study design
associated infections (HAIs) as a significant public health
problem, mainly due to their high impact on patients’ mor- This health economic evaluation study focused on financial
bidity, mortality and quality of life, but also due to their eco- data related to the SNM collaborative project. The Consolidated
nomic implications for health services globally [1]. Although Health Economic Evaluation Reporting Standards (CHEERS) 2022
HAIs have decreased in recent years [2,3], they still represent [15] (Table S1, see online supplementary material) and the
one of the most common adverse events related to patient care Principles of Good Practice for Budget Impact Analysis [16] were
[4,5]. For example, in the USA, it is estimated that approx- used as a basis, as recommended by the Professional Society for
imately 1.7 million patients are affected by HAIs each year, and Health Economics and Outcomes Research. Limitations, and
that these infections are responsible for 99,000 deaths, costing barriers to following these reporting standards are discussed
hospitals between US$35.7 and US$45 billion [1]. A recent below.
model developed by the National Health Service (NHS) in Eng-
land estimated an incidence of 653,000 HAIs among 13.8 million Health economic analysis plan: historical comparison
adult inpatients during the 2016e2017 period, which incurred a
cost of £2.1 billion [6]. All participating institutions were invited to report the
The five most significant HAIs e central-line-associated aggregate cost of HAIs and the number of patients with and
bloodstream infections (CLABSIs), ventilator-associated pneu- without HAIs in two defined periods: pre-intervention (Januar-
monia (VAP), surgical site infections, Clostridioides difficile yeMarch 2018) and post-intervention (AprileDecember 2018). A
infection and catheter-associated urinary tract infections form was sent to all institutions for completion, and partic-
(CAUTIs) e add an estimated US$9.8 billion annually in direct ipation was voluntary. Throughout the project, leaders
medical costs in the USA alone [7]. If the direct, indirect and employed ongoing strategies to generate engagement and
societal non-medical costs are combined, this figure ranges encourage institutions to report their financial data.
from US$96 to US$147 billion annually [8].
Diverse strategies exist to prevent and control HAIs, Study population
including surveillance programmes; hospital investigations in
the event of outbreaks; measures to prevent the spread of For the purposes of the SNM initiative, an open call was
infectious pathogens; and education for healthcare workers, made to all public institutions with adult ICUs throughout
patients and family members, [9]. The efficiency and effec- Brazil. Two hundred and twenty-six institutions applied for the
tiveness of such programmes must be considered to decide SNM project. Application was voluntary, and the participating
which should be implemented. Economic evaluations can organizations had no financial incentive to avoid conflicts of
determine which prevention strategies are cost-effective and interest. As pre-defined by the project governance, 120 ICUs
provide reasonable monetary value for hospitals, both indi- participated in the SNM activities; four withdrawals were
vidually and for the health system as a whole [10]. Even so, reported until December 2020. More detailed information on QI
measuring the cost of HAIs is still challenging, and economic methodology, patient care and relevant clinical indicators is
studies are often limited to simple value analyses, and need to available in previous publications [14,17].
adhere to reporting standards [11,12]. The inclusion criteria for the financial analysis module were:
Within the framework of the National HAI Prevention and (i) active participation in the SNM interventions; (ii) having a
Control Program and the National Patient Safety Policy, in cost and expense accounting system segregated by cost centre;
2017, the Brazilian Ministry of Health promoted a large-scale (iii) performing allocation of variable costs (supplies and drugs)
patient safety improvement initiative ‘Saúde em Nossas for each patient; and (iv) having a database containing infor-
Mãos’ (SNM; ‘Health in our hands’), which aimed to reduce HAIs mation about the patients treated by unit. These data were
in intensive care units (ICUs) using the Breakthrough Series organized into a pre-structured template following a 10-step
Collaborative (BTS) quality improvement (QI) model [13]. This process (Figure 1). The sole exclusion criterion was lack of
initiative reduced the incidence densities of CLABSIs, VAP and the necessary infrastructure to provide the information
CAUTIs by 43%, 52% and 68% from baseline, respectively, in 116 required by these 10 steps.
public ICUs [14]. This article aims to present the economic Forty-one institutions (35.6%) volunteered and completed
impact of SNM implementation, estimating the financial savings the template; however, only 13 (11.2%) provided the
and overall gain for Brazil’s Unified Health System (SUS). requested data with 100% accuracy.
10 R.M.C. Oliveira et al. / Journal of Hospital Infection 143 (2024) 8e17

Step 1 Step 2
ICU admission and discharge HAI confirmed? if yes: Inform HAI
Start (data and time – hour and minute). type and date of infection onset.
Responsibility of: Responsibility of:
Bed management service. Hospital infection control committee.

Step 4 Step 3
Material and medication consumption report. Medical prescription (tests, medications, materials,
Responsibility of: and procedures, including dialysis and transfusion).
Pharmacist. Responsibility of:
Medical team.

Step 5 Step 6
Tests (laboratory and imaging) utilization Procedures (medical procedures, dialysis,
report. and transfusions) utilization report.
Responsibility of: Responsibility of:
Medical diagnostics service. Medical care service.

Step 8 Step 7
Main procedure. Hospitalization Primary ICD-10 code for hospitalization.
authorization billing. Responsibility of:
Responsibility of: Bed management service.
Hospital management.

Step 9 Step 10
Number of official beds. Cost accounting for ICU expenses.
CNES Responsibility of: Responsibility of: Finish
National register of health Accounting service.
establishments or internal stats.

Figure 1. The 10 steps for completing the financial template. ICU, intensive care unit; HAI, healthcare-associated infection; ICD-10,
International Classification of Diseases, 10th revision.

Setting and location completion in December 2020, totalling 36 months of inter-

vention. A 3-month financial baseline between January 2018
The SNM initiative was promoted through the Support Pro- and March 2018 was established as a basis for further
gram for the Institutional Development of the Unified Health calculations.
System (PROADI-SUS); all participating ICUs were from public
hospitals. The SNM included ICUs from all five macro-regions of Data preparation
Brazil (North, North-east, Centre-West, South-east and South).
SUS is the world’s largest fully universal public healthcare The participating institutions filled out the template forms
system by several measures, including number of persons and were pre-assessed prior to further analysis. The PROADI-
covered, catchment area, and corresponding network size/ SUS financial team carefully verified the data received, using
number of facilities [18]. Regarding the local economic con- a detailed internal checklist to detect conformity between
text, by 2019, Brazil had an annual gross domestic product the care, procedures, tests, materials and medicines dis-
(GDP) of R$8.7 trillion (approximately US$2.2 trillion), and the pensed directly to the patient and the reported HAIs. In cases
GDP per capita was R$35,161 (approximately US$8,900) [19]. of inconsistencies, the form was returned to the source hos-
According to the National Treasury, in 2018, public health pital for adjustments until the data were 100% consistent.
expenditure corresponded to approximately 3.8% of GDP, close
to the 3.6% average of Latin American countries but below the Selection of outcomes
6.5% average of Organization for Economic Cooperation and
Development countries [20]. After receiving the completed forms with financial infor-
mation, the fixed cost (FC), variable cost (VC) and standard
Perspective and time horizon cost (SC) of the reported medical care were extracted for each
participating ICU. The FC refers to the expenses independent
The financial savings and overall gain were calculated from of the number of patients and care provided, and depends on
the SUS perspective. The time horizon was determined by the ICU capacity (i.e. wages, depreciation, inputs, supplies, com-
SNM duration period: from January 2018 to its planned monly used medications, and maintenance costs of the ICU).
 R.M.C. Oliveira et al. / Journal of Hospital Infection 143 (2024) 8e17 11

(a) (b)
Information sources
TDABC

Attended patients
Capacity cost rate
- Admission identification
= Cost of capacity supplied
- Data/time of ICU admission
- Identification of ICU admission with HAI Practical capacity of resources supplied

Unit cost
Fixed cost
- Resources allocation at the ICU (physical and human resources) = Cost of activity
- Segregation of cost and expenses by cost centre = Capacity cost rate x unit times
- Number of available beds for capacity calculation

Unit time
Variable cost
= Actual time needed to do the activity
- Resources used per patient
(Materials, drugs, laboratory, etc.)

Figure 2. (A) Sources for cost estimations. (B) Time-driven activity-based costing (TDABC). ICU, intensive care unit; HAI, healthcare-
associated infection.

The VC refers to those expenses which depend directly on the Measurement and valuation of the outcome: financial
medical care provided to each patient (Figure 2A). With these impact of the project
data on hand, the cost of each patient was estimated (¼FC þ
VC). The 13 financial spreadsheets gathered to calculate the
Next, the SC was calculated using the following formula: aggregated AC for each analysed HAI were used to assess the
The number of ICU admissions per month, patients with and financial impact of the SNM project.
without HAIs per month, and mean length of stay (LOS, in days) Infections that the project could have prevented were
were also recorded. The average cost (AC) for each analysed HAI estimated as reported elsewhere [14,17]. Thus, the aggregated
(CLABSIs, VAP and CAUTIs) per ICU was calculated as follows: AC from the pre-intervention period and the number of HAIs

Fixed cost ðall patientsÞ þ Variable cost ðall patientsÞ

Standard cost ¼
Total ICU admissions during the period of analysis

prevented by the initiative were used to demonstrate the

financial impact of the SNM project using the following
formula:

Fixed cost ðpatient with HAIÞ þ Variable cost ðpatient with HAIÞ
Average cost ðHAIÞ ¼
Total ICU admissions with HAI

Savings ¼ HAI aggregated AC  number of HAIs prevented in all participating ICUs until December 2020
12 R.M.C. Oliveira et al. / Journal of Hospital Infection 143 (2024) 8e17
The project investment throughout the PROADI-SUS is also beds). All institutions had an electronic hospital management
presented to determine the savings/gain margin. system and a financial invoicing system. Further features of the
participating ICUs are described in Table I.
Measurement and valuation of resources and costs The results of the groups with and without HAIs in the pre-
and post-intervention periods are shown in Table II. Regard-
The absorption model (also known as ‘full costing’) was used ing the pre-intervention period, cases of HAI amounted to 5.4%
to compile all costs related to the analysed HAIs, including (N¼171) of the total number of admitted patients (N¼3128),
direct costs (e.g. materials, human resources, drugs) and with higher LOS and AC when compared with the group without
indirect costs (e.g. bed-days, energy costs, management HAIs (3.4 and 3.9 times, respectively). On the other hand,
resources), in accordance with the recommendations of the during the post-intervention period, 10,103 patients were
Brazilian Ministry of Health [21]. Additionally, time-driven evaluated, and only 472 (4.9%) had HAIs. The indicators fol-
activity-based costing was used regarding the expected varia- lowed the same behaviour observed in the previous period: the
bility among institutions (Figure 2B) [22]. HAI group had a 3.2-fold longer LOS and a 3.4-fold higher AC
The SNM spreadsheet has been anonymized, codified and than the group without HAIs. The results of the post-
made available in the online supplementary material for fur- intervention period show a reduction of 0.9 percentage
ther details about costs and calculations. points in the number of admissions with HAIs, with a reduction
of 6% in LOS and 11% in AC.
Currency, price date and conversion The 13 institutions that completed the model could use their
HAI-aggregated AC for further calculations: R$43,677.11
For the purposes of this report, all values are presented in (US$8480) for CLABSIs, R$51,702.41 (US$10,039) for VAP, and
Brazilian reais (R$), followed by conversion to US dollars (US$) R$38,441.54 (US$7464) for CAUTIs. These values were used for
using the mean annual value of the corresponding year: 2020 further calculations for ICUs that did not apply or did not meet
[US$1 ¼ R$5.15 (minimum 4.02, maximum 5.93)] [23]. the minimum requested criterion. One ICU was excluded from
this economic evaluation due to incomplete data.
Ethics Finally, the overall savings/gain of the SNM project was
estimated through a projection using the aggregated AC for
Access to the SNM database was approved by the local each HAI, and the estimation of HAIs prevented for all 115
human research ethics committees (Certificate of Ethical participating ICUs until December 2020: 1727, 3797 and 2150
Appreciation 39657220.8.0000.5330) of the five PROADI-SUS CLABSIs, VAP and CAUTIs, respectively. This calculation resul-
institutions, with the consent of the SNM coordinator and the ted in an estimated economic impact of R$354,393,744 million
appropriate ministerial authorization. The available database (approximately US$68.7 million) to SUS (Table III).
presented financial data alone, and did not include any data The PROADI-SUS investment for the SNM project was
referring to or mentioning the participating institutions or the R$40,992,208 (approximately US$ 7.9 million), leading to an
patients involved. estimated return on investment (ROI) of 765% (R$7.6 per R$1.0
invested).
Results
Discussion
Of the 13 participating ICUs, 12 were from Southern and
South-eastern Brazil, and one was from the Northern region. The SNM initiative is one of the largest QI collaboratives
The institution sizes ranged from 115 to 743 beds (12e116 ICU worldwide, and one of the first in Latin America to use the BTS

Table I
Overview of the 13 institutions included in the cost analysis model of the ‘Saúde em Nossas Mãos’ initiative
Institution Region State Beds Management Computerized Financial EEC
All ICU system invoicing
system
1 SE Rio de Janeiro 148 50 Federal Yes Yes No
2 S Santa Catarina 202 30 SO Yes Yes Yes
3 SE Sao Paulo 316 44 State Yes Yes Yes
4 S Rio Grande do Sul 255 29 Federal Yes Yes Yes
5 SE Sao Paulo 743 116 State Yes Yes Yes
6 SE Sao Paulo 186 27 Other Yes Yes Yes
7 S Rio Grande do Sul 202 12 Federal Yes Yes Yes
8 SE Sao Paulo 297 76 SO Yes Yes No
9 N Pará 115 38 SO Yes Yes No
10 SE Minas Gerais 510 30 Federal Yes No Yes
11 SE Espı́rito Santo 189 20 Educational Yes Yes Yes
12 S Rio Grande do Sul 240 20 State Yes Yes Yes
13 SE Sao Paulo 202 21 State Yes Yes No
SE, South-eastern; S, Southern; N, Northern; SO, social organization; EEC, electronic extraction chart; ICU, intensive care unit.
 R.M.C. Oliveira et al. / Journal of Hospital Infection 143 (2024) 8e17 13
model for HAI prevention, significantly reducing CLABSIs, VAP
Analysis of admissions with and without healthcare-associated infections (HAIs) in the 13 participating intensive care units (ICUs) during the pre-intervention (JanuaryeMarch 2018)

R$41,773
US$8111
and CAUTIs in critical care settings [14]. The results of this

102,922
With

42,151

28,959
57,270
53,097
31,619
59,448
26,779
31,676
31,050
53,004
22,324
19,345
HAIs
Post-intervention
project led to a striking impact on the public health system in
terms of prevention, relevant clinical indicators, healthcare
quality, safety, and significant reduction of financial waste.
This report presents a reasonable and feasible approach to
Average cost per ICU (R$)

R$12,023
US$2334
Without

10,412 estimate the aggregated AC for each HAI analysed and the
14,826

25,755
19,389

27,526

19,933
16,410
HAIs

projection of savings among all participating ICUs, estab-

8821

6718

6988

8686
5436
4675
lishing the full financial impact of the nationwide project
from the perspective of a publicly funded health system.
The progressive increase in healthcare costs has become a

R$46,751
US$9077
significant challenge and a constant threat to accessing
116,284

104,696
With

30,210
59,881
45,648
34,351

36,730
29,598
30,224
45,501
28,785
16,358
HAIs

health systems worldwide. WHO has revealed that, in many

Pre-intervention

countries, the total health expenditure exceeds GDP [24]. To

0

ensure optimal efficiency of the financial resources provided

to healthcare institutions, proper management of the costs

R$11,885
involved in patient care is mandatory [25]. Such management

US$2307
Without

15,342
16,425

25,605
14,096

30,262

15,477
11,783
HAIs

is particularly critical in low- and middle-income countries,

8108

7549

8041

8609
4963
5007 where obtaining maximum efficiency with chronically low
budgets is crucial to providing reasonable monetary value in
programmes to control HAIs [26e31].
Despite a wealth of evidence demonstrating the impor-
With
HAIs
19
14
15
17
18
17
12
15
15
14
25
14
17
16
Post-intervention

tance of cost analysis for decision-making, calculating the cost

of HAIs is still challenging [32,33]. A review from 2005 assessing
70 published studies on the cost of HAIs in hospitals found that
Mean length of stay (days)

the economic evidence on HAI control efforts needed to be

Without
HAIs

more compelling because of the variety of study designs and

5
4
5
8
8
4
5
4
9
7
5
3
4
5

settings, statistical methods and cost outcomes used [34].

Likewise, another systematic audit of economic evidence
linking HAIs and preventive measures from 1990 to 2000, also
performed by Stone et al. [35], and a recent systematic review
With
HAIs
0
15
18
18
21
20
18
19
14
14
22
18
15
17
Pre-intervention

by Arefian et al. [36] found that economic evaluation guide-

lines were not followed; therefore, the quality of the reports,
according to the Professional Society for Health Economics and
Outcomes Research, was low.
Without
HAIs

Despite knowledge of reporting standards, the current

6
3
5
8
6
5
5
4
8
6
6
3
5
5

financial team found several barriers to following the inter-

national recommendations, and a full assessment of economic
repercussions from all participating ICUs could not be com-
pleted. As evidenced herein, most institutions needed help to
With
HAIs

118

115

472
6

4
15
22
24

12
32
17

51

22
34
Post-intervention

inform the authors of the resources that patients used during

their ICU stays. As a result, only 13 hospitals could be included
and post-intervention (April 2018eDecember 2018) periods
Patients admitted per month (mean)

in this economic analysis. Inadequately detailed hospital cost

data e needed to accurately reflect the economic value of the
Without

resources used to control each type of HAI e as well as the

1206
1102

1060

1210

9631
562

625
352

453
590
271
860
834

506
HAIs

broad heterogeneity of systems in use, lack of internal align-

ment (non-centralized) and, consequently, lack of stand-
ardization of the costing process hampered further comparison
among all participating ICUs. Moreover, some institutions have
With
HAIs

171
0

3
14
16
27

14
11

12
43

11
12

more than one type of computerized system and no interface

Pre-intervention

among these, so data were not recorded and integrated

properly. In most ICUs participating in this study, adjustment of
routine data collection methods was necessary, which led to
Without

strain and delays. These adjustments were done manually with

2957
111
326
340
221
138
319
131
171
111
275
267
395
152
HAIs

the support of an expert team.

Furthermore, Brazil has one of the world’s most unequal
income distributions [18,37]. All these inequities directly
influence healthcare institutions across the country, including
Institution

differences in infrastructure and human resources, financial

gaps, and the individual characteristics of different services,
Table II

Total

which represent limitations for uniform financial data report-

10
11
12
13
1
2
3
4
5
6
7
8
9

ing, as observed.
14 R.M.C. Oliveira et al. / Journal of Hospital Infection 143 (2024) 8e17

Table III
Estimated savings of the ‘Saúde em Nossas Mãos’ initiative
Healthcare-associated infections Aggregated average Infections Savings (R$) (min/max)
cost (R$) (min/max) prevented (N)
Central-line-associated 43,677.11 (24,029/185,059) 1727 75,430,376 (41,498,306/319,598,547)
bloodstream infections
Ventilator-associated 51,702.41 (17,963/128,708) 3797 196,314,049 (68,208,046/488,706,892)
pneumonia
Catheter-associated 38,441.54 (17,638/95,788) 2150 82,649,319 (37,923,789/205,945,885)
urinary tract infections
Total R$354,393,744 (147,630,142/1,014,251,325)
US$68,814,319 (28,666,047/196,942,004)

Most prominent economic analyses and financial models activities are incorporated into the whole and are not detailed
have been conducted in North American and European coun- per activity. The value of these strategies should be e whenever
tries [34e36], which seems unsuitable as a point of reference possible e standardized, to provide a better vision of specific
for making decisions about the Brazilian healthcare system. In expenses and inform planning of future initiatives. The costs of
addition, the methods used to adjust the financial assessments collaborations need to be factored into the translation of
between countries need to be studied in greater depth before improvements, and their costs or cost-effectiveness need to be
their practical implementation in foreign healthcare systems evaluated to identify savings in healthcare budgets and benefits
[38]. It follows that economic evaluations of HAI control pro- to society [44e47]. Overall, due to the challenges mentioned
grammes in local or regional settings for developing countries above, there is a need to highlight the difficulties of conducting a
are desirable [26]. Cost estimation of HAIs in critical care set- high-quality economic assessment and of following international
tings has been reported previously in Brazil, evidencing direct reporting guidelines, which can be exhausting and time-
costs (up to an eight-fold increase) and increased LOS; in fact, consuming, especially when analyses are conducted across
LOS attributed to HAIs has the most significant financial impact institutions with different levels of financial organization.
on the public health system [39e41], which is in line with the Simply reducing costs without considering the results is a
present findings and reflects the significant negative impact of significant threat to the quality of care [48]. QI initiatives
HAIs for SUS. These results represent a benchmark for further should simultaneously consider three aspects to ensure value-
economic analysis of HAI prevention strategies in Brazilian based health and make the health system more efficient e
settings. improved population health, the care experience, and reduced
Another leading challenge is financial estimation in critical costs per capita e as any change in one can affect the others
care settings, where seriously ill patients will already have [49]. In light of this, SNM improved patient outcomes and
received high-cost care regardless of their infection status healthcare quality indicators [14,17] and reduced costs for the
[42]. Additionally, collateral costs related to suboptimal healthcare system. These results confirmed that SNM is a fea-
treatment, adverse drug reactions and the emergence of sible and profitable approach for the Brazilian healthcare sys-
antibiotic resistance, among others, should also be considered tem, and a value-based strategy for preventing HAIs in critical
in financial calculations [43], without mentioning the indirect care settings.
and non-medical societal costs related to these infections [8]. One potential gain observed in this study was an increase of
Given these considerations, defining the most accurate values 2.8 percentage points in the availability of ICU beds (i.e.
for the input variables is extremely challenging. These studies patients with HAIs occupied 12.6% of beds before the inter-
are even more difficult to perform with traditional economic vention vs 9.8% after the intervention). This suggests that by
evaluation models, as hospital care is a non-standard proce- reducing HAIs, bed occupancy could be optimized without
dure (e.g. patients with the same diagnosis may require dif- additional investments in the existing infrastructure. A recent
ferent combinations of tests and treatments due to different NHS model shows a cost of £$2.1 billion in 2016e2017 related to
comorbidity profiles). HAIs [6], 99.8% of which was attributable to patient manage-
A particular challenge for the SNM initiative was the value ment, resulting in 5.6 million occupied hospital bed days
estimation of QI strategies implemented, such as face-to-face attributable to HAIs. Hospital inpatients who acquire an HAI
training sessions, leadership huddles, virtual meetings, feed- have a higher likelihood of extending their LOS. Hence, the
back and process improvement methods, among many others costs associated with HAIs may be primarily attributable to
[44]. All these activities require planning, management and in- patients’ increased LOS, which has also been reported in low-
practice application, as well as time-consuming coordination, [50,51], middle- (including Brazil) [30,40,52] and high-income
travel, materials and variable expenses, which subjectively settings [53e55].
increase the cost of QI programmes [45,46]. A 2018 systematic Hand hygiene is an essential measure to reduce the spread
review identified 64 QI strategies reporting effectiveness meas- of infection, and it was one of the main goals of the SNM pro-
ures in clinical care outcomes and, consequently, savings; how- gramme, as reported previously [14,17]. The economic impli-
ever, only four studies reported cost-effectiveness outcomes for cations of successful hand hygiene improvement have long
these collaborations [46]. To fulfil this demand, the total cost been established. It has been demonstrated that these pro-
reported by the PROADI-SUS institutions is presented, but QI grammes cost less than 1% of the HAI-related costs, making
 R.M.C. Oliveira et al. / Journal of Hospital Infection 143 (2024) 8e17 15
them relatively inexpensive and unequivocally worth the areas, analysis, and dissemination of costs. This study confirms
investment [56]. Good hand hygiene programmes also depend the importance of analysing the economic impact of HAI pre-
on systemic components of the healthcare facility, including vention programmes, suggesting that a QI model to prevent
access to supplies and how nurses are allocated, organized and HAIs in critical care settings is a feasible value-based approach,
trained. The work culture is crucial for implementing QI reducing financial waste and having a significant ROI for the
strategies to improve adherence, achieve better clinical out- publicly funded Brazilian healthcare system.
comes, and generate savings [28].
Acknowledgements
Limitations
The authors wish to thank all the healthcare professionals
This evaluation has some limitations. First, the cost analysis working in the participating ICUs, and the PROADI-SUS techni-
proposed a pre- and post-intervention design. In academic cal and administrative teams who supported the development
terms, it would be desirable to have a control group to assess of this project.
the financial impact of the implementation when compared
with usual care or another multi-modal HAI control strategy. Conflict of interest statement
Second, as mentioned before, estimating the real-life added None declared.
value of HAIs in intensive care settings is difficult. Several
factors e such as the inherently critical condition of ICU Funding sources
patients, healthcare worker preferences and experiences, This study was conducted with public funding from the
hospital features, and the specificities of different health Ministry of Health through PROADI-SUS, and with charitable
services, among others e may influence HAI AC in each par- funding from the participating institutions: Hospital Alemão
ticipating ICU. Further studies are needed to enhance the Oswaldo Cruz, Hcor, Hospital Israelita Albert Einstein, Hos-
accuracy of HAI added-cost estimation, including demographic pital Moinhos de Vento, and Hospital Sı́rio-Libanês. None of
and severity scores. the institutions were involved in the design and writing of
Third, cost-effectiveness and cost-utility analyses, includ- this publication.
ing a societal perspective, are advisable in any health eco-
nomics analysis and would help improve understanding of the
financial impact of the SNM initiative; however, these analyses
Appendix A. Supplementary data
were not initially proposed by the project governance and were
Supplementary data to this article can be found online at
not included in the initiative.
https://doi.org/10.1016/j.jhin.2023.10.001.
Finally, regarding the difficulties, barriers and challenges
discussed throughout this report, it was not possible to carry
out a full financial assessment of all participating ICUs, and the References
small number of ICUs for which it was possible to complete such
an analysis (N¼13) may not be representative of the sample, [1] World Health Organization. Guidelines on core components of
let alone the country as a whole. However, a reasonable and infection prevention and control programmes at the national and
acute health care facility level. Geneva: WHO; 2016. Available at:
feasible approach for savings estimations has been presented,
https://apps.who.int/iris/handle/10665/251730.
which can serve as a basis for benchmarking in Brazil, dem- [2] Johnson NB, Hayes LD, Brown K, Hoo EC, Ethier KA; Centers for
onstrating the impact of a universal, unified, publicly funded Disease Control and Prevention (CDC). CDC National Health
healthcare system eSUS e rather than local savings across the Report: leading causes of morbidity and mortality and associated
Brazilian regions. Additionally, the range (minimum and max- behavioral risk and protective factors e United States, 2005e
imum) of estimated savings has been reported to represent the 2013. MMWR Suppl 2014;63:3e27.
many possible scenarios, considering divergences among ICUs. [3] Magill SS, O’Leary E, Janelle SJ, Thompson DL, Dumyati G,
Nadle J, et al. Changes in prevalence of health care-associated
The future infections in U.S. hospitals. N Engl J Med 2018;379:1732e44.
[4] Rosenthal VD, Duszynska W, Ider BE, Gurskis V, Al-Ruzzieh MA,
Myatra SN, et al. International Nosocomial Infection Control
These findings lead to a better understanding of HAI control.
Consortium (INICC) report, data summary of 45 countries for
They may serve as a basis for further studies and prevention 2013e2018, adult and pediatric units, device-associated module.
projects for a wider audience, and are essential to informing Am J Infect Control 2021;49:1267e74.
more appropriate and cost-effective public health policies. [5] Allegranzi B, Bagheri Nejad S, Combescure C, Graafmans W,
The project is still ongoing in Brazil, and new ICUs are being Attar H, Donaldson L, et al. Burden of endemic health-care-
targeted, with the hope of expanding the success of this ini- associated infection in developing countries: systematic review
tiative across the country. This study was the first to estimate and meta-analysis. Lancet 2011;377:228e41.
the project’s overall savings, serving as an experience to [6] Guest JF, Keating T, Gould D, Wigglesworth N. Modelling the
improve the quality of economic reports by following interna- annual NHS costs and outcomes attributable to healthcare-
tional guidelines. associated infections in England. BMJ Open 2020;10:e033367.
[7] Zimlichman E, Henderson D, Tamir O, Franz C, Song P, Yamin CK,
In conclusion, economic evaluation of HAIs remains chal-
et al. Health care-associated infections: a meta-analysis of costs
lenging, mainly when analyses are performed in institutions and financial impact on the US health care system. JAMA Intern
with diverse levels of financial organization. There is a pressing Med 2013;173:2039e46.
unmet need to organize, standardize and systematize the [8] Marchetti A, Rossiter R. Economic burden of healthcare-
costing process in healthcare systems at all stages of admis- associated infection in US acute care hospitals: societal per-
sion, from data recording, interfaces between systems and spective. J Med Econ 2013;16:1399e404.
16 R.M.C. Oliveira et al. / Journal of Hospital Infection 143 (2024) 8e17
[9] Bryant KA, Harris AD, Gould CV, Humphreys E, Lundstrom T, Health/PAHO; 2013. Available at: https://bvsms.saude.gov.br/
Murphy DM, et al. Necessary infrastructure of infection pre- bvs/publicacoes/introducao_gestao_custos_saude.pdf.
vention and healthcare epidemiology programs: a review. Infect [26] Gamalathge PU, Kularatna S, Carter HE, Senanayake S, Graves N.
Control Hosp Epidemiol 2016;37:371e80. Cost-effectiveness of interventions to reduce the risk of
[10] Stone PW. Economic burden of healthcare-associated infections: healthcare-acquired infections in middle-income countries: a
an American perspective. Expert Rev Pharmacoecon Outcomes systematic review. J Infect Prev 2019;20:266e73.
Res 2009;9:417e22. [27] Singh S, Kumar RK, Sundaram KR, Kanjilal B, Nair P. Improving
[11] Rennert-May E, Conly J, Leal J, Smith S, Manns B. Economic outcomes and reducing costs by modular training in infection
evaluations and their use in infection prevention and control: a control in a resource-limited setting. Int J Qual Health Care
narrative review. Antimicrob Resist Infect Control 2018;7:31. 2012;24:641e8.
[12] Stone PW, Larson E, Kawar LN. A systematic audit of economic [28] Thi Anh Thu L, Thi Hong Thoa V, Thi Van Trang D, Thuy Van D,
evidence linking nosocomial infections and infection control Phuc Tien N, Thi Kim Anh L, et al. Cost-effectiveness of a hand
interventions: 1990e2000. Am J Infect Control 2002;30:145e52. hygiene program on health care-associated infections in intensive
[13] Institute for Healthcare Improvement. The breakthrough series: care patients at a tertiary care hospital in Vietnam. Am J Infect
IHI’s collaborative model for achieving breakthrough improve- Control 2015;43:e93e9.
ment. Cambridge, MA: IHI; 2003. Available at: http://www.ihi. [29] Rosenthal VD, Udwadia FE, Kumar S, Poojary A, Sankar R,
org/resources/Pages/IHIWhitePapers/ Orellano PW, et al. Clinical impact and cost-effectiveness of
TheBreakthroughSeriesIHIsCollaborativeModelforAchieving split-septum and single-use prefilled flushing device vs 3-way
BreakthroughImprovement.aspx. stopcock on central line-associated bloodstream infection rates
[14] Tuma P, Vieira Junior JM, Ribas E, Silva KCCD, Gushken AKF, in India: a randomised clinical trial conducted by the Interna-
Torelly EMS, et al. A national implementation project to prevent tional Nosocomial Infection Control Consortium (INICC). Am J
healthcare-associated infections in intensive care units: a col- Infect Control 2015;43:1040e5.
laborative initiative using the Breakthrough Series Model. Open [30] Rattanaumpawan P, Thamlikitkul V. Epidemiology and economic
Forum Infect Dis 2023;10:ofad129. impact of health care-associated infections and cost-
[15] Husereau D, Drummond M, Augustovski F, de Bekker-Grob E, effectiveness of infection control measures at a Thai university
Briggs AH, Carswell C, et al. Consolidated Health Economic hospital. Am J Infect Control 2017;45:145e50.
Evaluation Reporting Standards 2022 (CHEERS 2022) Statement: [31] Luangasanatip N, Hongsuwan M, Lubell Y, Limmathurotsakul D,
updated reporting guidance for health economic evaluations. Srisamang P, Day NPJ, et al. Cost-effectiveness of interventions
Value Health 2022;25:3e9. to improve hand hygiene in healthcare workers in middle-income
[16] Sullivan SD, Mauskopf JA, Augustovski F, Jaime Caro J, Lee KM, hospital settings: a model-based analysis. J Hosp Infect
Minchin M, et al. Budget impact analysis e principles of good 2018;100:165e75.
practice: report of the ISPOR 2012 Budget Impact Analysis Good [32] Scott 2nd RD, Culler SD, Rask KJ. Understanding the economic
Practice II Task Force. Value Health 2014;17:5e14. impact of health care-associated infections: a cost perspective
[17] Tuma P, Vieira Júnior JM, Ribas E, Silva KCCD, Gushken AKF, analysis. J Infus Nurs 2019;42:61e9.
Torelly EMS, et al. The impact of the coronavirus disease 2019 [33] Graves N, Harbarth S, Beyersmann J, Barnett A, Halton K,
(COVID-19) pandemic on a national project preventing Cooper B. Estimating the cost of health care-associated infec-
healthcare-associated infections in intensive care units. Infect tions: mind your p’s and q’s. Clin Infect Dis 2010;50:1017e21.
Control Hosp Epidemiol 2023. https://doi.org/10.1017/ [34] Stone PW, Braccia D, Larson E. Systematic review of economic
ice.2023.146. analyses of health care-associated infections. Am J Infect Control
[18] Paim J, Travassos C, Almeida C, Bahia L, Macinko J. The Brazilian 2005;33:501e9.
health system: history, advances, and challenges. Lancet [35] Stone PW, Larson E, Kawar LN. A systematic audit of economic
2011;377:1778e97. evidence linking nosocomial infections and infection control
[19] Brazilian Institute of Geography and Statistics. Painel de Indica- interventions: 1990e2000. Am J Infect Control 2002;30:145e52.
dores. Rio de Janeiro: IBGE; 2023. Available at: https://www. [36] Arefian H, Vogel M, Kwetkat A, Hartmann M. Economic evaluation
ibge.gov.br/indicadores.html. of interventions for prevention of hospital acquired infections: a
[20] National Treasury Department, Brazil. Aspectos Fiscais da Saúde systematic review. PLoS One 2016;11:e0146381.
no Brasil. Brası́lia: National Treasury Department; 2018. Available [37] Stalker P. A guide to countries of the world. 3rd ed. Oxford:
at: https://www.tesourotransparente.gov.br/publicacoes/ Oxford University Press; 2016.
aspectos-fiscais-da-saude-no-brasil/2018/30. [38] Vieira SF. Evidências econômicas de intervenções em saúde sob a
[21] Ministry of Health, Brazil. Programa Nacional de Gestão de Cus- perspectiva do sistema único de saúde: por que e para que
tos. Manual Técnico de Custos: conceitos e metodologia. Brası́lia: produzi-las e utilizá-las? J Bras Econ Saúde 2017;9:229e36.
Ministry of Health; 2006. Available at: https://bvsms.saude.gov. [39] Osme SF, Almeida APS, Lemes MF, Barbosa WO, Arantes A,
br/bvs/publicacoes/06_0243_M.pdf. Mendes-Rodrigues C, et al. Costs of healthcare-associated
[22] Cabral HD, Neto V, de Souza RL. Time-driven activity-based costing infections to the Brazilian public unified health system in a
(TDABC): uma revisão sistemática da literatura. In: Proceedings of tertiary-care teaching hospital: a matched caseecontrol study.
the Brazilian Congress of Costs e ABC; 2021. Available at: https:// J Hosp Infect 2020;106:303e10.
anaiscbc.emnuvens.com.br/anais/article/view/4866. [40] Osme SF, Souza JM, Osme IT, Almeida APS, Arantes A, Mendes-
[23] Central Bank, Brazil. Cotações e boletins. Central Bank; 2020. Rodrigues C, et al. Financial impact of healthcare-associated
Available at: https://www.bcb.gov.br/estabilidadefinanceira/ infections on intensive care units estimated for fifty Brazilian
historicocotacoes. university hospitals affiliated to the unified health system. J Hosp
[24] World Health Organization. Public spending on health: a closer Infect 2021;117:96e102.
look at global trends. Geneva: WHO; 2018. Available at: https:// [41] Leal MA, Freitas-Vilela AA. Costs of healthcare-associated infec-
www.who.int/publications/i/item/WHO-HIS-HGF- tions in an intensive care unit. Rev Bras Enferm
HFWorkingPaper-18.3. 2021;74:e20200275.
[25] Ministry of Health, Brazil/Pan American Health Organization. [42] Mylonakis E, Ziakas PD. How should economic analyses inform
Introdução à Gestão de Custos em Saúde. Brası́lia: Ministry of nosocomial infection control? AMA J Ethics 2021;23:E631e8.
 R.M.C. Oliveira et al. / Journal of Hospital Infection 143 (2024) 8e17 17
[43] Ziakas PD. Economic considerations of antimicrobial stewardship patients admitted to a tertiary hospital in Ethiopia: longitudinal
programs. In: Antimicrobial stewardship: principles and practice. study. Antimicrob Resist Infect Control 2018;7:2.
2016. p. 322e33. [51] Gidey K, Gidey MT, Hailu BY, Gebreamlak ZB, Niriayo YL. Clinical
[44] de la Perrelle L, Radisic G, Cations M, Kaambwa B, Barbery G, and economic burden of healthcare-associated infections: a
Laver K. Costs and economic evaluations of quality improvement prospective cohort study. PLoS One 2023;18:e0282141.
collaboratives in healthcare: a systematic review. BMC Health [52] Jia H, Li L, Li W, Hou T, Ma H, Yang Y, et al. Impact of healthcare-
Serv Res 2020;20:155. associated infections on length of stay: a study in 68 hospitals in
[45] Schouten LM, Hulscher ME, van Everdingen JJ, Huijsman R, China. Biomed Res Int 2019;2019:2590563.
Grol RP. Evidence for the impact of quality improvement col- [53] Benenson S, Cohen MJ, Schwartz C, Revva M, Moses AE, Levin PD.
laboratives: systematic review. BMJ 2008;336:1491e4. Is it financially beneficial for hospitals to prevent nosocomial
[46] Wells S, Tamir O, Gray J, Naidoo D, Bekhit M, Goldmann D. Are infections? BMC Health Serv Res 2020;20:653.
quality improvement collaboratives effective? A systematic [54] Manoukian S, Stewart S, Dancer S, Graves N, Mason H,
review. BMJ Qual Saf 2018;27:226e40. McFarland A, et al. Estimating excess length of stay due to
[47] Carter AW, Mandavia R, Mayer E, Marti J, Mossialos E, Darzi A. healthcare-associated infections: a systematic review and meta-
Systematic review of economic analyses in patient safety: a analysis of statistical methodology. J Hosp Infect
protocol designed to measure development in the scope and 2018;100:222e35.
quality of evidence. BMJ Open 2017;7:e017089. [55] Manoukian S, Stewart S, Graves N, Mason H, Robertson C,
[48] Porter ME. What is value in health care? N Engl J Med Kennedy S, et al. Bed-days and costs associated with the inpa-
2010;363:2477e81. tient burden of healthcare-associated infection in the UK. J Hosp
[49] Berwick DM, Nolan TW, Whittington J. The triple aim: care, Infect 2021;114:43e50.
health, and cost. Health Aff 2008;27:759e69. [56] Pittet D, Sax H, Hugonnet S, Harbarth S. Cost implications of
[50] Ali S, Birhane M, Bekele S, Kibru G, Teshager L, Yilma Y, et al. successful hand hygiene promotion. Infect Control Hosp Epi-
Healthcare associated infection and its risk factors among demiol 2004;25:264e6.