Evaluation of quality indicators of cervical cytopathology tests carried out in a municipality of Paraná, Brazil

Introduction: Although the criteria and recommendations are well defined to improve the quality of cervical cytopathology tests (Pap smear) in Brazil, the quality indicators of several laboratories are below the recommended standards. Objective: To evaluate the quality indicators of cervical cytopathology tests carried out by the Brazilian Unified Health System [Sistema Único de Saúde (SUS)]. Method: Cross-sectional study based on the results of cervical cytopathology tests conducted in the municipality of Cascavel, Paraná, Brazil, from January 2012 to December 2018. The study included 141,191 tests and the following variables were analyzed: sample suitability, cervical cytopathology tests result, and internal quality monitoring indicators recommended by the Brazilian Ministry of Health. Results: The percentage of unsatisfactory tests did not exceed 0.97%. The representation of the squamocolumnar junction ranged from 63% to 67.4%. The percentage of altered exams among the satisfactory samples ranged from 0.46% to 6.44%. The positivity index ranged from 0.46% to 6.44%, while the high-grade squamous intraepithelial lesion/satisfactory indicator ranged from 0.21% to 1.06%, atypical squamous cells of undetermined significance/altered ranged from 22.47% to 49.93%, atypical squamous cells of undetermined significance/satisfactory ranged from 0.1% to 2.57%: and atypical squamous cells of undetermined significance/squamous intraepithelial lesions ranged from 0.31 to 1.02. Conclusion: In 2012 and 2013, the high-grade squamous intraepithelial lesion/satisfactory and the positivity index indicators were below the recommended parameters, and in the remaining years, these indicators were within the parameters recommended. Consequently, a greater number of truly cancer precursor lesions were identified, demonstrating the importance of internal quality control in the prevention of cervical cancer.

When early diagnosed, CC can reach 100% cure. In this regard, the cervical cytopathology test, when performed with quality, is the test of choice for detecting this cancer in Brazil (3) . Variables in the pre-analytical phase and in the analytical phase of cervical cytopathology testing, such as representation of the epithelium in the collection, correct fixation, identification of the material, cytologist's experience level, appropriate working conditions, implementation of standard operating procedures, in addition to qualification and periodic training of professionals, may impair their sensitivity (4,5) .
In cytopathology, quality control is based on the detection, correction and reduction of deficiency in the execution proceedings within the laboratory, resulting in cytopathology tests with greater quality and reliability, minimizing diagnostic errors as much as possible and collaborating to improve the preanalytical and analytical phases, in addition to contributing to public health (6) .
In Brazil, the quality control of cytopathology tests was recommended from 2001 onwards through Ordinance SPS/SAS no. 92 of the Brazilian Ministry of Health (MH) (7) . This ordinance was ratified by Ordinance MH no. 3.388 of December 30, 2013, which instituted the Brazilian national qualification in cytopathology in preventing cervical cancer [Qualificação nacional em citopatologia na prevenção do câncer do colo do útero (Qualicito)], which consists of the definition of standards and criteria in the evaluation of the quality of the cytopathology test by monitoring public and private Brazilian Unified Health System [Sistema Único de Saúde (SUS)]. service providers, classified as type I and type II laboratories. Type I laboratories are public and private laboratories that provide services to SUS and perform cytopathology tests; type II are the public laboratories responsible for performing cytopathology tests within the scope of External Quality Monitoring (EQM) (8) .
The MH/Inca prepared the Quality Management Manual for Cytopathology Laboratories, aiming to guide the SUS laboratories service provider on the implementation of Internal Quality Monitoring (IQM) and EQM (6) . The Manual and Qualicito encourage increased coverage of the target audience (women aged 25 to 64 years), and the monitoring of internal quality indicators promotes permanent education of health professionals, encourages and establishes parameters for SUS service providers' contracting and cancelation of agreement (6)(7)(8) .
The continuous monitoring of the results is an important action that should be performed by the cytopathology laboratory, based on the estimated indicators found in the Quality Management Manual to assess the professionals' global and individual performance (6) . Some of these indicators are: positivity index (PI), percentage of tests compatible with high-grade squamous intraepithelial lesions (HSIL) in satisfactory tests, percentage of atypical squamous cells (ASC) of undetermined significance in altered and/or satisfactory tests and squamous atypia's of undetermined significance and squamous intraepithelial lesions ratio (6) .
Despite these efforts, studies show results that are not recommended by the MH, such as the percentage of tests performed outside the recommended age range (9) . In the state of Minas Gerais, Brazil, laboratories providing services to the SUS presented PI below the recommended by the MH, since several laboratories did not detect or detected a low number of CC precursor lesions (10) ; another study results show that the quality indicators for laboratories that provide services to SUS in several states and regions of Brazil are, for the most part, outside the parameters recommended by the MH (11,12) .
Although criteria and recommendations are well defined for improving the quality of cytopathology test in Brazil, the quality indicators of several laboratories are below the standards recommended by the MH. Therefore, the objective of this study was to evaluate the quality indicators of cervical cytopathology tests performed by SUS in a municipality in the state of Paraná state, Brazil. The variables analyzed were: sample suitability, result of the cytopathology test, and internal quality monitoring indicators.
Information on the results of cytopathology tests, including the suitability of the sample and the representativeness of the squamocolumnar junction (SCJ) were extracted from the cytopathological reports contained in SISCOLO and SISCAN and classified according to the Brazilian nomenclature for cervical reports and recommended conduct (13) .
The evaluation of the indicators for internal monitoring of the quality of the cytopathology test was carried out based on five indicators, which are described below with their respective formulas, as recommended by the MH (14) : a) PI -PI expresses the prevalence of cellular changes and characterizes the sensitivity of screening to detect lesions in the population examined, according to the Brazilian nomenclature for cervical reports and recommended conduct (14) .

Calculation method:
Total cytopathology tests with altered results in a specific place and period × 100 Total satisfactory cytopathology tests performed in the same place and period For the critical analysis of laboratories registered in SISCOLO/ SISCAN, the PI categorization was determined as follows: very low -below 2%; low -between 2% and 2.9%; expected -between e2282020 Evaluation of quality indicators of cervical cytopathology tests carried out in a municipality of Paraná, Brazil 3% and 10%; above the expected -above 10%. It is necessary to consider that these providers can assist secondary referral services for cervical pathology. b) percentage of tests compatible with HSIL/satisfactory tests: HSIL represent the CC truly precursor lesions, that is, those that actually have the potential for progression. Its detection is the main objective of secondary prevention of CC. It is recommended that the indicated value is greater than or equal to 0.4%.
Calculation method:

Number of HSIL tests × 100
Total satisfactory tests c) percentage of ASC/altered tests -PI must be analyzed together with the ASC/altered tests indicator, as this index, apparently adequate, may contain a high percentage of tests compatible with ASC. It is recommended that the indicated value is below 60% of the altered tests. Calculation method:

Number of tests with ASC-US and ASC-H × 100
Total satisfactory tests e) ASC/squamous intraepithelial lesions (SIL) ratio -assists in the identification of alteration that are low-grade squamous intraepithelial lesions (LSIL) and HSIL. It is recommended that this ratio should not exceed 34.
Calculation method:

Number of ASC-US and ASC-H tests Number of tests with LSIL and HSIL
Data analysis was performed using descriptive statistics (calculation of absolute and relative frequencies), using the Microsoft Excel software.

results
The results of this study showed that about 15,859 and 22,546 tests were performed per year; the percentage of unsatisfactory tests did not exceed 0.97%. Over the years, the SCJ representation varied from 63% to 67.4% of the tests ( Table 1).
The total of satisfactory samples ranged from 15,701 to 22,147 tests per year, the largest number of altered tests in 2017, with 1,364 cytopathological results. Among the altered exams, the results that presented ASC-US ranged from 0.08% to 2.22%; those with ASC-H, from 0.02% to 0.53%; those with LSIL, from 0.12% to 2.58%; and those with HSIL, from 0.21% to 1.06%. From 2015 to 2018, results of invasive carcinoma were also identified ( Table 2).
The results show that PI ranged from 0.46% to 6.44% (value within the recommended parameter, between 3% and 10%) in five of the seven years of study. Regarding the HSIL/satisfactory tests indicator, the index ranged from 0.21% to 1.06% -value within the recommended (≥ 0.4%) in five from the seven years of study. Regarding the ASC/altered tests indicator, the index ranged from 22.47% to 49.93% -a value within the recommended range (< 60%) in all years of the study. The ASC/satisfactory exams indicator ranged from 0.1% to 2.57%, a value within the recommended range (< 5%) in all years of the study. Finally, the ASC/SIL ratio indicator ranged from 0.31 to 1.02, a value within the recommended range (< 3) in all years of the study ( Table 3).

discussiOn
The results of this study show that, over the years, there has been an increase in the detection of HSIL and HSIL with suspicious invasion characteristics (HSIL micro). The high-grade lesion is considered the true precursor lesion of the CC; its detection and treatment are important to avoid evolution to cancer (15) .
The increase in HSIL detection directly impacts the HSIL/ satisfactory tests and PI indicators. The HSIL/satisfactory tests indicator measures the ability to detect precursor lesions, and the PI, the prevalence of cellular changes in tests, as well as the sensitivity of the screening process to detect lesions in the examined population. A low PI can indicate that malignant changes are not being identified, which causes false-negative   e2282020 Evaluation of quality indicators of cervical cytopathology tests carried out in a municipality of Paraná, Brazil results (6) . In this study, we noticed that, as of 2014, these indicators were within the recommended parameters, differently from the years 2012 and 2013, whose values were much lower than the recommended. We can infer that this scenario occurred due to the strict internal quality control in 2014, which was implemented in the laboratory service provider for SUS, which contributed to the standardization of cytomorphological criteria and, consequently, to the improvement in quality indicators.
The ASC/altered tests, ASC/satisfactory tests, and ASC/SIL indicators were within the recommended parameters in all years of the study. The ASC/altered tests indicators should be analyzed together with the PI, as an apparently suitable PI contains a high percentage of ASC. A large number of borderline atypia may point to problems, both in sample collection and/or preparation and in the cytopathological analysis. Overall, this demonstrates the need for professionals continuing education for, especially for the review of ASC diagnostic criteria (6) .
The year 2013 was the one with the lowest detection of malignant changes and, consequently, the lowest rates. In part, this can be justified by the lower representation of SCJ in the analyzed samples. The SCJ representation in cytological smears is very important, as it is recognized as the region most likely to develop CC precursor lesions (16) .
Less severe changes, ASC-US and LSIL, were the most frequent, except in 2013. ASC-US are correlated with low severity disease for most women, so the expected clinical management is conservative (17) , as with low-grade lesion, which has a significant spontaneous regression rate (18) and is considered the least likely lesion to progress to invasive carcinoma (19) .
Regarding the sample suitability, the number of unsatisfactory samples did not exceed the 5% recommended by the MH (6) and its main causes were acellular or hypocellular material and desiccation artifacts. Among the rejected samples, the main reason was identification absence or error. These findings corroborate those found by Galvão et al. (2015) (20) : from the unsatisfactory samples, 34% were due to acellular or hypocellular material; 21% were rejected due to identification absence or error; and 14% showed desiccation artifacts.
The quality of the cytopathology test is closely linked to factors such as information on anamnesis data, collection performed properly, good smear fixation, adequate color and careful analysis of the sample. Any mistake in one of these factors may lead to false-negative or false-positive results and harm not only women, but society as a whole.
In order to improve the quality control of the cytopathology test, the MH, in 2013, published ordinance 3,388 called Qualicito (8) . Among the recommendations of the ordinance, as a quality criterion, the laboratories may have an annual production of 15,000 tests, which was demonstrated in this study.
As limitations of this work, we highlight the lack of information on the population profile. This data would contribute to explain the prevalence of less severe changes and other possible variables that influence the PI.

cOnclusiOn
The results of this study show that, in 2012 and 2013, the HSIL/satisfactory tests and PI indicators were below the parameters recommended by the MH. In the remaining years, these indicators were within the recommended parameter. Consequently, a greater number of CC truly precursor lesions was identified. Therefore, we demonstrate the importance of implementing internal quality control of cytopathology tests for CC screening and prevention.