With the increasing use of health economic evaluation (HEE) in decision-making and health resource allocation and management policy design has seen an increase in HEE studies on screening programs in China l. In addition to the quantity of HEE, the quality may be of particular concern as it influences the reliability of HEE evidence adopted in policy formulation. This study sought to assess the reporting quality of HEE on screening programs over the last 20 years in China and identify potential predictors and relevant recommendations to improve the quality of study reporting.

A search of HEE studies published in PubMed, Embase, CNKI and WANFANG from 2000 to 2021 was performed. Two reviewers independently extracted data and assessed the quality if reporting using the 24 item Consolidated Health Economic Evaluation Reporting Standards ( CHEERS) checklist. The CHEERS score for each study was converted into standardized 0-1 point scale. General liner regression was used to identify predictors associated with the reporting quality.

One hundred and thirty-three studies met the inclusion criteria. The mean standardized score for the included studies was 0.56 (title), 0.64 (abstract), 0.74 (introduction), 0.58 (methods), 0.40 (results), 0.70 (discussion), and 0.54 (other section). The number and reporting quality of articles published each year showed an overall upward trend. A greater proportion of studies were published in Chinese journal (69.2%), modelling-based (54.9%), conducted by universities/research institutions (45.9%), focused on non-infectious disease (84.2%), using cost-effectiveness analysis method (50.4%), published in non-specialty journal (60.2%), and declaring the funding support (76.7%). Items related to study perspective, discount rate, measurement of effectiveness, currency and price, analytical methods, uncertainty, heterogeneity and conflicts were under-reported. Published year, journal type, first author affiliation and economic evaluation type predicted higher score in regression analyses (p<0.05).

Overall, the quantity and quality of HEE on screening programs in China is improving, although there is a need to improve the use of on specific reporting items in the CHEERS criteria. The use of suitable evaluation guidelines will make the decision-making process more scientific.

Diagnostic testing and patient monitoring are important to diagnose potential diseases and to evaluate treatment regimens. Since diagnosis and treatment monitoring have no intrinsic effects, an economic evaluation of biomarkers is inevitably linked to the resulting therapeutic interventions, which depend on both physicians’ decisions and diagnostic accuracy of the test (i.e., sensitivity, specificity). In this review we analyzed the methodology of economic studies evaluating the management of the five most relevant non-communicable diseases, that is, obesity, cancer, diabetes, cardiovascular and chronic respiratory diseases.

A systematic search in Medline and the National Health Service Economic Evaluation Database (NHS EED) covering the last ten years served to identify health economic analyses of biomarkers used in diagnosing / monitoring. Findings were reviewed with respect to analytical method, reported outcomes and comparability.

The search yielded 680 abstracts in total out of which 280 full texts were reviewed and 77 sources included following predefined criteria. Most economic analyses (94%) evaluated the clinical outcome and costs of testing / monitoring in correlation to a corresponding intervention, 6 percent of the sources focused on the accuracy of the test or monitoring methods only. There were 61 studies that included an economic model; overall, 15 sources presented the outcome as cost per life year gained (CEA), 37 sources as cost per QALY gained (CUA), and 12 provided the outcome as both a CEA and CUA. In 16 analyses the outcome was presented in other economic terms as, for example, cost per additional case detected.

Determining the value of biomarkers requires consideration of the clinical consequences of a test result (incorrect treatment decisions, impact on morbidity, mortality, or quality of life) as well as the corresponding economic outcomes. Most of the identified studies considered at least one of these aspects. Results are presented in manifold ways but do not necessarily address decision makers’ needs. Thus, clear guidelines on economic evaluations of biomarkers are needed and should include broader health system views like affordability or the number of unnecessary interventions avoided.

The public health areas without commercial value continue to be underserved, while those of high profit for industry will not be sustainable for much longer. We hypothesize that the lack of equity and efficiency in the biomedical research and development system is mainly due to a pharma-led short-term profit orientation that ignores the values of other relevant stakeholders.

This pilot study reached some consensus on the principles of a co-created biomedical research and development process based on the preferred supplier (PS) model, which proposes a public health procurement system prioritizing business with companies fulfilling the “4 Share” criteria of priorities, risks and rewards, results, and outcomes to ensure that health needs are met.

A constructive health technology assessment, which included VALues In Doing Assessments of health TEchnologies (VALIDATE) methodology, was used to analyze the values and dissent of a pilot sample of ten global key informants. The methodology comprised qualitative techniques such as an online preliminary survey, in-depth semi-structured interviews, and a Delphi survey to reach a joint construction by reconstructing the stakeholders’ interpretive frames and applying an adaptation of the Richardson model to contested values.

There was consensus on combining efficiency and social justice norms by incentivizing diseases affected by market failure due to small subpopulations (e.g., rare diseases), low availability to pay, restricted use (e.g., antibiotics), and difficulty demonstrating results (e.g., Alzheimer’s disease). Stakeholders mainly agreed on the PS 4 Share principles, highlighting the need for price to be linked to impact modulated by tracked research and development costs and investments, as proposed by the PS model. More market incentives such as push, and especially pull incentives (market access), should be included. The PS model should be cause-solution oriented, promote open-disruptive innovation, and guarantee fast patient access.

The PS model achieved multistakeholder consensus for desirability (values) and plausibility (shifting incentives). Stakeholders pointed out the need to apply health technology assessment to further develop the model and pilot it in the European Union with the NextGenerationEU.

Horizon scanning (HS) is an early health technology assessment (HTA) method for raising early awareness among policymakers of promising new and emerging health technologies. The Agency for Care Effectiveness (ACE) in Singapore established a HS system in 2019 to complement its HTA process. While the HS system initially focused on cell and gene therapies, this was expanded in 2020 to include medical technologies (MTs). This abstract focuses on the role of HS in nominating MTs for early evaluation to guide subsidy decisions, with the intention of avoiding challenges in altering deeply entrenched practices.

The ACE methodology for HS aligns with the core principles and methods of international best practice. Generally, MTs addressing national healthcare priorities are tracked. To identify topics for subsidy evaluation, the local registration status of an MT was used as the main selection criterion because of its proximity to the technology’s early diffusion into the healthcare system. MTs with regulatory approval were selected for HTA and subsidy consideration. All nominated technologies were checked against eligibility criteria for HTA and then assessed against a standard checklist for prioritizing HTA topics.

Among the 1,025 MTs tracked by the HS system, 89 were locally registered and nominated for HTA. Following eligibility assessment, 26 topics remained. After the prioritization exercise six topics were shortlisted. To date, two evaluations have been completed to guide subsidy decisions and four topics are undergoing evaluation. Notably, 16 of the 26 eligible topics were excluded due to a lack of sufficient evidence, in terms of both quantity and quality, for evaluation.

HS can be a useful tool for identifying new MTs for evaluation and possible funding prior to further diffusion, but careful selection of the technologies is required to ensure a sufficient evidence base for evaluation. Moving forward, HS can also play a more active role in disinvestment of obsolete or low value health technologies.

The number of anti-diabetics listed in the Ministry of Health Medicine Formulary (MoHMF) and the proposal to change prescribing category for usage in primary care facilities has increased each year causing difficulty in monitoring effectiveness, safety and rational use of the treatment and budget management. This study aimed to describe the reassessment and selection strategy of Dipeptidyl Peptidase-4 Inhibitors (DPP4-i), a class of anti-diabetic medicines in the MoHMF.

A literature review on the comparative effectiveness and safety of all available DPP4-i in MoHMF were conducted. Comparative treatment cost and utilization of DPP4-i were analyzed. Approved MoHMF indications were listed and compared against approved Drug Control Authority (Malaysia) indications. Information on clinical guidelines recommendations and listing or reimbursement status in non-Ministry of Health Malaysia (MoHM) public institutions and other countries were obtained. All findings were presented to MoHM drug expert committee for DPP4-i selection strategy. The final recommendation based on consensus among drug expert committee and Pharmaceutical Services Program were presented to the MoHMF Panel.

The MoHMF Panel acknowledged that the efficacy and safety profile of all DPP4-i were equivalent across therapeutic group as supported by strong evidence, hence, their selection can be made via cost-minimization strategy. The call for contract tender for single tablet was conducted based on therapeutic group (drug class) to encourage price competition and contracts were awarded to two DPP4-i which offered the lowest treatment costs. Saxagliptin and vildagliptin were awarded as contract items while sitagliptin remained as local purchase item. Prescribing category for all DPP4-i in MoHMF were streamlined accordingly. Linagliptin was disinvested due to sufficient availability of alternatives.

Selection strategy and disinvestment has successfully reduced the number of DPP4-i listed in MoHMF thus allowing more efficient clinical and cost monitoring. Cost minimization through tender by therapeutic group was the first to be done and has efficiently provided an avenue for price competition which results in saving to MoHM.

The National Institute for Health and Care Excellence (NICE) intends to automate the way it monitors the uptake, impact, and value of its guidance. Traditionally this has been done by developing impact reports, long documents that, while well received, are time consuming to develop and can quickly become outdated.

We focused on a novel topic that would benefit from new data sources to examine its impact: a rapid guideline for managing the long-term effects of coronavirus disease 2019 (COVID-19). We shortlisted “measurable” recommendations within the guideline that were likely to be captured in data collections. We then reviewed available data sources that included relevant up-to-date data. Finally, we explored what existing methods were available to NICE for automating impact reporting.

For long COVID-19 we accessed OpenSAFELY, a secure, transparent software platform for primary care COVID-19 data that was developed in response to the pandemic. This captured data on the management of long COVID-19 in primary care as well as onward referral to specialist clinics. In addition, we accessed data from the CVD-COVID-UK/COVID-IMPACT Consortium, which links general practice records with primary care dispensing data. This enabled us to analyze the impact of the pandemic on the prescribing and dispensing of cardiovascular disease medications. Working with our digital team we developed an automated impact reporting dashboard using Google’s data studio. This enabled different views of the data, for example by region or socioeconomic status, to be presented in an automated way.

Automating the impact reporting of NICE guidance provides up-to-date information on its value to the health system. While we were able to collect new sources of data and automate some aspects of how these were viewed, full automation requires several enablers. These include an application programming interface between the data sources and NICE, and ensuring that NICE guidance is computer readable so that its measurement is practical in healthcare systems.

Clinical trials that fail prematurely due to poor design are a waste of resources and deprives us of data for evaluating potentially effective interventions. This study used machine learning modelling to predict clinical trials’ success or failure and to understand feature contributions driving this result. Features to power the modelling were engineered using data collected from the National Institute for Health and Care Research Innovation Observatory’s ScanMedicine database.

Using ScanMedicine, a large dataset containing 641,079 clinical trial records from 11 global clinical trial registries, was extracted. Sixteen features were generated from the data based on fields relating to trial design and eligibility. Trials were labeled positive if they were completed (or target recruitment was achieved) or negative if terminated/withdrawn (or target recruitment was not achieved). To achieve optimal performance, phase-specific datasets were generated, and we focused on a subsample of Phase 2 trials (n=70,167). Ensemble models using bagging and boosting algorithms, including balanced random forest and extreme gradient boosting classifiers were used for training and evaluating predictive performance. Shapley Additive Explanations was used to explain the output of the best performing model and calculate feature contributions for individual studies.

We achieved a weighted F1-score of 0.88, Receiver Operator Characteristic Area under the Curve score of 0.75, and balanced accuracy of 0.75 on the test set with the xgBoost model. This result shows that the model can successfully distinguish between classes to predict if a trial will succeed or fail and subsequently output the features driving this outcome. The number of primary outcomes, whether the study was randomized, target sample size and number of exclusion criteria were the most important features affecting the model’s prediction.

This study is the first to use predictive modelling on a large sample of clinical trial data obtained from 11 international trial registries. The prediction outcomes achieved by our novel approach, which uses phase-specific trained models, outperforms previous modelling in this space.

The recent National Institute for Health and Care Excellence (NICE) consultation on methods for health technology evaluation (HTE) outlined plans to include guidance on the role of real-world evidence (RWE) in HTE. This is particularly pertinent for interventions where the evidence base consists of a single arm trial (SAT) design, which is more frequently observed in oncology. This study reviewed the influence of RWE submitted as part of the evidence base for NICE oncology appraisals.

A search for NICE HTEs was conducted for interventions supported by SATs from January 2017-November 2021. Evidence was stratified by submission packages with SAT evidence alone or in combination with randomized controlled trial (RCT) evidence, with or without RWE.

Thirty-two decisions for interventions supported by SATs were made by NICE between 2017-2021, all in oncology indications. Fifty percent were supported by SAT evidence and fifty percent by RCT plus SAT evidence, both with or without RWE. A lower proportion of RCT/ SAT HTEs submitted RWE compared to SAT HTEs (fifty vs ninety four percent). Seventy five percent and nineteen percent of SAT HTEs received a positive recommendation, with and without restrictions, irrespective of submitting RWE. One negative decision was observed for SATs supported by RWE. Sixty three percent and thirty eight percent of RCT/ SAT HTEs received a positive recommendation, with and without restrictions. Overall, the proportion of positive recommendations were lower for HTEs submitting RWE (ninety six percent) compared to HTEs not submitting RWE (one hundred percent), which is in contrast to recent findings specific to orphan oncology HTEs (one hundred versus seventy eight percent).

RWE was more commonly submitted to support SAT HTEs, than RCT HTEs. The use of RWE seems to be established as a necessity to supplement a SAT evidence base, whereas RWE is more generally a nice to have in RCT HTEs. However, RWE appears to positively influence decision-making for orphan oncology indications with a more neutral influence for non-orphan indications.

The Brazilian government has made efforts in systems to generate data from medical devices (MD). This work explores the main systems and data sources in the perspective of contributing as a source to generate real world data (RDW).

Document review of relevant national data sources for MD. In addition, a structured search was carried out in EMBASE using key descriptors for RWD applied to the regulatory context and to the management of health technologies, without date or language restrictions.

Eighteen primary federal government data sources for MD were identified. Not all sources are publicly accessible. Of the articles, the search returned 1,185 results, of which 29 titles were selected and 8 met the protocol’s objective. Included articles were from Europe, the United States and Canada. As in other countries, Brazil initially systematized DM administrative data to meet commercial and financial demands. With the evolution of health technology assessment methods, the use of RDW has become imperative to assess the value of MD to society. Common examples from these countries are implantable MD databases. Current challenges focus on data linkage and quality, in addition to standardized naming. The adoption of the Unique Device Identification (UDI) is one of the promising initiatives to facilitate traceability throughout the lifecycle proposed in the International Medical Device Regulators Forum (IMDRF) of which Brazil is a member. Among the systems, the following stand out: i) ConectSUS, which intends to provide access to health information centered on the patient, anywhere and at any time; ii) National implant registry that generates data on implanted prostheses and stents, surgical techniques used, the profile of patients and the health services involved.

This work showed the similarities between Brazil and other countries in the management of MD data throughout its life cycle, as well as mapped the national primary data sources for MD.

Prostate cancer was the second most frequent cancer and the fifth leading cause of cancer death among men in 2020. The incidence rates vary substantially in countries with different Human Development Indexes (HDI), while the mortality rates decrease with improved access to the health system, availability of therapies and earlier detection. Worldwide, population-based cancer registries are important tools for planning and managing health systems. The Fundação Oncocentro de São Paulo (FOSP) is responsible to collect, clean and publicize data from cancer treatment institutions. This study aimed to describe retrospectively the demographic and clinical profile of prostate cancer (PC) in Brazil using this database. It is not an incidence study as data is representative only from specific institutions.

This was a retrospective observational study of the years 2000 to 2020 from analysis of the publicly available FOSP database (http://www.fosp.saude.sp.gov.br).

The records were extracted, merged, and cleaned using a fully documented and validated data process. Only patients included on the register with a primary PC diagnosis were considered.

From January 2010 to June 2020, there were 943,660 patients diagnosed with C61 in FOSP database for the considered time period. The majority of the FOSP database records are from patients who live and/or were born in SP (91.8 and 58.4%, respectively) or MG (2.8 and 10.5%, respectively). The mean age of PC at baseline was 69 years. Considering the stage of the disease, the mean ages are 55, 70, 67, 66 and 61, for stages 0, 1, 2, 3 and 4, respectively. This cohort was also analyzed in relation to treatments received, and status at the end of treatment (51.3% are disease-free, 18.4% are alive with cancer, and 30.3% are dead).

FOSP population-based cancer registries are a powerful tool to obtain information for planning, and improving the management of healthcare services especially for São Paulo.

The National Institute for Health and Care Excellence (NICE) intends to increasingly use real-world evidence in developing guidance. To increase trust in such evidence, NICE has developed a framework for developing and assessing real-world evidence studies, including understanding the value of the selected data source for the decision problem.

Starting with published high-quality studies about data quality, we developed a conceptual model of the elements needed to understand the quality of a data source. Results from a literature search were then mapped to the model. We used this to design a structured reporting tool, the data Suitability Assessment Tool (dataSAT), and tested it in several cases studies. Additionally, we engaged with internal and external stakeholders to obtain feedback on the tool and revised it accordingly.

DataSAT covers provenance of the data, assessment of data quality, and the data’s relevance to the research question. For data provenance, information is requested about the data source independent of the study’s interests, including the purpose, setting, dates of operation, funding, data specification, and management and quality assurance plans for the data sources. Data quality is covered by quantitively assessing the completeness and accuracy of the following key study elements to inform critical appraisal of the study: population inclusion and exclusion criteria; intervention; comparator; and outcomes and key covariates. The findings on data sources and data quality are then interpreted in terms of relevance to the decision problem. This includes relevance to the population in the United Kingdom, the treatment pathway and care setting, the availability of key study elements, time-related factors such as length of follow up, and the effects of sample size and missing data on the validity of findings.

DataSAT allows summary information on source data, including quality and relevance, to be reported in a structured manner, enabling decision makers to better understand how the data influence the robustness of analyses used in health technology assessment. This helps increase trust in the use of real-world evidence.

Recent reviews have shown that many real-world evidence (RWE) studies suffer from avoidable methodological flaws. Meanwhile, the National Institute for Health and Care Excellence (NICE) is seeing an increase in RWE submissions in Health Technology Appraisals and is keen to support the use of this evidence. However, limited guidance exists for the development and assessment of RWE, risking both missed opportunities for unbiased evidence generation and inconsistent decision making based on that evidence. As part of its RWE framework, NICE has developed methods guidance to provide clear expectations for the conduct and reporting of non-randomized comparative effects studies using real world data.

A conceptual model and draft framework were developed based on established international best practices in RWE and observational research. This was refined with focused literature searches, for example, on the use of external control arm studies. We then engaged with external stakeholders to incorporate their feedback and develop case studies. A reporting template was developed and tested on multiple use cases.

The guidance stresses the central importance of a target trial approach to study design, e.g., adopting an active comparator, new user design, where possible. Target trial emulation is a useful tool to improve the quality and transparency of RWE studies, helping to overcome selection and confounding biases. Various other study design and analytical approaches are outlined for addressing confounding bias and biases due to missing data, measurement error, or misclassification, which are common challenges in RWE. Alongside traditional approaches to sensitivity analysis, the framework promotes quantitative bias analyses which includes a range of methods to assess and communicate the potential impact of remaining bias to study findings by quantifying the direction, magnitude, and uncertainty of bias. A reporting template, based on common methodological pitfalls, is provided to help evidence developers consider key areas of bias in their work and to inform reviewers of any approaches used to investigate or resolve these.

Randomized controlled trials (RCTs) are typically considered the gold standard source of clinical evidence for reimbursement submissions, but they can often be resource-intensive, expensive, and may not always be appropriate. For example, it may be unethical to assign patients to an untreated or undiagnosed control group, or blinding may not be feasible when assessing medical devices. Evidence for medical devices is therefore often limited to nonrandomized studies. We explored the use and value of real-world evidence (RWE) in the reimbursement of medical devices across several health technology assessment (HTA) agencies.

A narrative review was completed to compare the acceptability of RWE for the HTA evaluation of medical devices across a convenience sample of countries. English-language published guidance documents were reviewed, and study design preferences extracted.

In Australia, France, Germany, Ireland, Norway, and Scotland, HTA agencies prefer RCT evidence but accept RWE as supporting data. In England, there is no preferred study design, with directly observed clinical outcomes, evidence syntheses, nonclinical, and modelling studies accepted. Notably, methods and processes for HTA programs are being reviewed and are expected to place a greater emphasis on RWE. In Australia, pseudo-randomized trials, comparative cohort studies, case series, and other study designs are permitted. In France, nonrandomized or nonblinded trials, patient preference cohorts, prospective comparative observational studies, and propensity score matched cohorts are permitted, accompanied by justification. In Scotland, lived experiences, RWE, and systematic reviews are accepted. In Germany, nonrandomized studies are deemed to provide “minimum”, “very low” or “low” certainty of results. In Norway, RWE may be accepted if no RCT data are available, or to support RCTs.

In the assessment of medical devices, where RCTs are unsuitable, RWE can form a feasible alternative. Real-world evidence is increasingly being recognized as a valuable source of evidence for medical interventions and is accepted by a number of HTA agencies. No funding was received for this study.

Appropriate resource utilization is crucial for cancer screening programs. Overscreening is defined as screening provided beyond the upper age limit of the target age or at a shorter interval than recommended in national programs. In Japan, there are no upper age limits set for cancer screening programs, and the recommended screening interval for cervical and breast cancer screening is 2 years. To examine the efficient use of resources for cervical and breast cancer screening, we investigated how often overscreening occurred in both programs.

The target age for this study was defined as 20-69 years for cervical cancer screening and 40-69 years for breast cancer screening. We used the national report for cancer screening in 2017 in Japan and estimated the number of participants over 70 years old or those who participated in screening annually. The percentage of overscreening was compared between cervical cancer and breast cancer screening by chi-square test.

The number of participants was 4,294,127 for cervical cancer screening and 3,087,781 for breast cancer screening in 2017. The percentage of overscreening in total participants was 38.0 percent for cervical cancer screening and 35.7 percent for breast cancer screening (p<0.01). The percentage of screening at overage was higher in breast cancer screening than in cervical cancer screening (21.1% vs. 13.9%, p<0.01), whereas more frequent screening was seen more often in cervical cancer screening than in breast cancer screening (29.7% vs. 19.6%, p<0.01). If the resources used in overscreening could be used for the target population, it was estimated that the participation rate could increase by 4.1% for cervical cancer screening and 4.3% for breast cancer screening.

In Japan, screening for overage participants and short intervals may have contributed to unnecessary screening for cervical cancer and breast cancer. These resources used for overscreening could be allocated to screening for the target population.

Hepatocellular cancer (HCC) is a severe condition with poor prognosis and a significant burden. Selective internal radiation therapy (SIRT) is recommended as an alternative treatment option to overcome some limitations of current treatments. This analysis estimated the cost-effectiveness of SIRT using Y-90 resin microspheres for the treatment of unresectable HCC in Brazil.

This study was conducted from the Brazilian payer perspective according to local guidelines, with a lifetime horizon. The use of SIRT in patients with intermediate- or advanced-stage HCC, without extrahepatic disease and ineligible to transarterial chemoembolization, was compared with sorafenib, the commonly used HCC systemic treatment in Brazil. A sensitivity analysis included the subgroup of patients with low tumour burden and preserved liver function.

A partitioned-survival model was developed, which included a tunnel state for patients downstaged to receive a treatment with a curative intent such as liver surgery, transplantation or ablation. Survival curves, utilities and adverse events incidence were extracted from published sources of pivotal randomized control trials. Effectiveness of health interventions was measured in quality-adjusted-life-years (QALYs) and life-years (LYs). Local costs from Brazil were applied. Future costs and effects were discounted at five percent. A willingness-to-pay threshold of USD 53,936 was used, based on a 2017 review of healthcare technology adoption in Brazil.

LYs and QALYs were higher for SIRT using Y-90 resin microspheres versus sorafenib (0.27 and 0.20 incremental LYs and QALYs, respectively) and costs were slightly higher for SIRT (USD 3,056 incremental costs). The incremental cost-effectiveness ratio (ICER) was USD 14,948 per QALY in the basecase.

One-way and probabilistic sensitivity analyses confirmed the robustness of the analyses. Scenario analyses tested different model assumptions and reinforced the basecase results indicating that SIRT using Y-90 resin microspheres was highly likely to be cost-effective compared with sorafenib. Also, the ICER was lower in the subgroup compared with the overall population.

SIRT using Y-90 resin microspheres represents a cost-effective option compared with sorafenib in Brazil.

Therapeutic drug monitoring (TDM) is a cost-effective tool to increase treatments’ efficacy and safety. Analyses of pharmacokinetics proprieties of tyrosine kinase inhibitors (TKI) used for chronic myeloid leukemia (CML) can contribute towards effective TDM, development of tailored treatments and new dosing regimens. We aimed to synthesize the available evidence on pharmacokinetic parameters of imatinib, nilotinib, bosutinib, ponatinib in healthy individuals vs. CML patients.

Systematic searches were conducted in PubMed, Scopus and Web of Science. We included in vivo studies addressing TKIs’ pharmacokinetics, including maximum observed concentration (Cmax), time of maximum observed concentration (Tmax) and half-life (t1/2). Meta-analyses of event rates (mixed-effect models) were performed for the parameters of interest: area under the concentration-time curve from time zero to the last measurable concentration (AUC0-t) and from time zero to infinity (AUC 0-∞). Results were presented as event rates with 95 percent confidence intervals. Heterogeneity was assessed using chi-square and I2 statistical tests and considered significant when p<0.05 and high when I2>75 percent (Comprehensive Meta-Analysis v.2 Biostat-Englewood, NJ).

Overall, 50 articles were included for analyses (n=26 imatinib, n=11 nilotinib, n=8 bosutinib, n=5 ponatinib). Most studies were performed in the United States (46.0%), designed as open-label trials (70.0%). Several significant interactions between TKI with enzyme inhibitors (ketoconazole, midazolam, aprepitant, metoprolol, grapefruit juice), proton pump inhibitors (esomeprazole, lanzoprazole, omeprazole), antacids, H2 antagonists (famotidine) and enzyme inducers (St. John’s, rifampicin) were found (p<0.001). Given the significant increase in AUC and Cmax in patients with hepatic/liver impairments currently using TKI, strict therapeutic monitoring is paramount to maintain safety. The between study heterogeneity was rated as moderate to high (I2=75-90%) due the limited number of trials for some drugs, different study design, and populations.

The co-administration of TKI with hepatic enzyme inducers or inhibitors, proton pump inhibitors, antacids, H2 antagonists, as well as in patients with hepatic/liver failures requires caution and additional monitoring. Further well-designed trials are needed to strengthen this evidence for some TKIs, namely bosutinib and ponatinib.

Non-palpable breast cancers require intraoperative localization to guide the surgical procedure. The radio-guided occult lesion localization (ROLL) and radioactive seed localization (RSL) techniques use radioactive material (technetium-99m labeled macroaggregated albumin and iodine-125 seeds, respectively) implanted at the lesion site. The success of conservative surgery depends on complete tumor excision with negative surgical margins. The objective of this study was to perform a meta-analysis of the surgical effectiveness of the ROLL and RSL techniques with respect to rates of positive surgical margins, reoperation, and recurrence.

A systematic search of the PubMed, Embase, LILACS, SciELO, and Web of Science databases was performed on 2 August 2021. A standard form was used to extract data from the studies selected in the screening process. The ROBINS-I tool was used to analyze risk of bias and RevMan 5.4 software was used to perform the meta-analysis.

Four observational studies met the inclusion criteria. The overall evidence quality was moderate. The studies included women aged from 28 to 91 years with non-palpable breast cancer. The results demonstrated similar effectiveness between RSL and ROLL for rates of positive surgical margins (risk ratio [RR] 0.83, 95% confidence interval [CI]: 0.50, 1.39; 763 patients) and reoperation (RR 1.14, 95% CI: 0.75, 1.74; 1,550 patients). Regarding the rate of disease recurrence, RSL was superior to ROLL (RR 0.50, 95% CI: 0.29, 0.87; 939 patients).

The results demonstrate that the ROLL and RSL techniques are equivalent with respect to rates of positive surgical margins and reoperation, although patients undergoing RSL had lower rates of disease recurrence. However, there is a tendency to favor the RSL technique because of the longer interval between implantation and surgery, which is possible due to the longer half-life of iodine-125 (59.4 days). This also means that radioactive seed implantation can occur before neoadjuvant chemotherapy, so the tumor bed remains marked if further interventions are required, obviating the need for another invasive procedure before surgery.

There were an estimated 55,102 new cases of colorectal cancer (CRC) in Brazil in 2020, comprising 9.3 percent of all newly diagnosed cancers, and making it the third most common cancer in the Brazilian population. Up to half of all patients with CRC will develop liver metastases, and the prognosis for these patients is poor. The objective of the present analysis was to evaluate the cost-utility of selective internal radiation therapy (SIRT) using Y-90 resin microspheres versus best supportive care (BSC) in patients with unresectable, liver-dominant, chemotherapy-refractory metastatic CRC (mCRC).

A three-state partitioned survival model was developed in Microsoft Excel to evaluate the cost-utility of SIRT using Y-90 resin microspheres versus BSC. Membership of the three model states of pre-progression, post-progression and death was governed by parametric models of Kaplan-Meier data from a retrospective, interventional study. Costs associated with SIRT using Y-90 resin microspheres, BSC, and adverse events were obtained from Brazilian sources and reported in 2021 United States dollars (USD). Future costs and effects were discounted at 5 percent. One-way and probabilistic sensitivity analyses (PSA) were performed. A willingness-to-pay threshold of USD 53,936 was used based on a 2017 review of Brazilian healthcare technology adoption.

The base case analysis showed that Y-90 resin microspheres would result in an increase of 0.76 QALYs versus BSC, increasing quality-adjusted life expectancy from 0.67 QALYs to 1.43 QALYs. The improvement in quality-adjusted life expectancy was accompanied by an increase in costs from USD 9,884 to USD 40,399 over the model time horizon, corresponding to an increase of USD 30,515, and yielding an incremental cost-utility ratio (ICUR) of USD 40,265 per QALY gained. PSA showed there was a 90.8 percent likelihood of cost-effectiveness at USD 53,936 per QALY gained.

SIRT using Y-90 resin microspheres was cost-effective versus BSC in the treatment of unresectable, liver-dominant mCRC patients and should be considered for incorporation in the Brazilian Private Healthcare System.

We aimed to develop and validate machine learning (ML) -based algorithms to predict COVID-19 diagnosis as well as to identify new biomarkers associated with the disease.

Initially, 96 blood samples of patients diagnosed with COVID-19 (Thaizhou Hospital, China) were analyzed through liquid chromatography coupled to mass spectrometry. Samples of patients presenting other pneumonias or severe acute respiratory syndrome, but with negative RT-PCR for SARS-CoV-2, were used as positive controls. Samples from healthy volunteers were used as negative controls. The final database included around 1000 metabolites. Exploratory analyses for the development of ML-based models using principal component analysis (PCA) were performed. Leverage plot versus studentized residuals method was used to detect outliers. Three supervised ML-based models were developed: discriminant analysis by partial least squares (PLS-DA), artificial neural networks discriminant analysis (ANNDA) and k-nearest neighbors (KNN). Samples for the training (70%) and testing sets (30%) were randomly selected using the Kenrad Stone algorithm. Models’ performance was evaluated considering accuracy, sensitivity and specificity. Analyses were conducted in SOLO (Eigenvector-Research).

The PCA model was able to distinguish the three classes of patients’ samples (positive for COVID-19, negative controls, positive controls) with an overall accumulated variance of 94.27 percent. The PLS-DA model presented the best performance (accuracy, sensitivity, and specificity of 93%, 98% and 88%, respectively). Increased levels of the biomarkers uridine (linked to glucose homeostasis, lipid, and amino acid metabolisms), 4-hydroxyphenylacetoylcarnitine (metabolite from the tyrosine metabolism; probably associated with anorexia) and ribothymidine (resulting from oral and fecal microbiota alterations) were significantly associated with COVID-19.

Three different and updated ML-based algorithms were developed to predict COVID-19 diagnosis; PLS-DA led to the most accurate results. High levels of some metabolites were found as potentially predictors of the disease. These biomarkers should be further evaluated as potential therapeutic targets in well-designed clinical trials. These ML-based models can help the early diagnosis of COVID-19 and guide the development of tailored interventions.

We aimed to map and synthesize the available evidence on neuron-specific biomarkers related to COVID-19.

A systematic review and qualitative evidence mapping synthesis was performed (PROSPERO-CRD42021266995). Searches were conducted in PubMed and Scopus, and complemented by manual search (July 2021). We included observational studies of any design assessing neurological biomarkers in adult patients (>18 years; with or without neurological comorbidities) diagnosed with COVID-19. Methodological quality of nonrandomized studies (case-control, cohorts) was assessed using the Newcastle-Ottawa Scale.

Overall, 14 studies (n=485 patients) conducted in Sweden (n=4 articles, 28.5%), Germany (n=3; 21.4%), USA (n=3; 21.4%), Canada, France, Italy and Norway (n=1 study each) were included. The most reported neurological symptoms (n=13 studies, 92.8%) were headache, confusion, general weakness, loss of smell/taste, cognitive impairments and behavioral changes. Prevalent neurological conditions included encephalopathies, neuropathies, myopathies, and delirium; most critical cases presented cerebrovascular events (n=4 studies, 28.5%). Hypertension, diabetes, obesity, dyslipidemia, and chronic lung disease were the most reported comorbidities. Eight different neuron-specific biomarkers were found in primary studies: neurofilament-light chain – NfL (n=10 studies; 71.4%), glial fibrillary acidic-protein – GFAp (n=5; 35.7%), tau protein (n=5; 35.7%), neurofilament-heavy chain – NfH, S100B protein, ubiquitin C-terminal hydrolase L1 - UCH-L1, neuron-specific enolase and beta protein-amyloid – Aβ (n=1 study each). These biomarkers were found both in cerebrospinal fluid and blood/plasma samples even without an evident cytokine storm. In patients with COVID-19, NfL and GFAp can act as sensitive indicators of neuroaxonal and astrocytic damages, respectively. Increased levels of NfL were significantly associated with severe COVID-19, unconsciousness and longer stay in the intensive care unit (p<0.05). Studies had an overall poor to moderate methodological quality.

We identified eight neuron-specific biomarkers that should be further studied as prognostic factors of COVID-19. These findings can also guide the development of targeted therapies against SARS-CoV-2. Additional well-designed clinical trials are needed to strengthen this evidence and help understand the mechanisms of neurological symptoms and sequelae after COVID-19 infection.