Cardinality estimation is a fundamental but long unresolved problem in query optimization. Recently, multiple papers from different research groups consistently report that learned models have the potential to replace existing cardinality estimators. In this paper, we ask a forward-thinking question: Are we ready to deploy these learned cardinality models in production? Our study consists of three main parts. Firstly, we focus on the static environment (i.e., no data updates) and compare five new learned methods with nine traditional methods on four real-world datasets under a unified workload setting. The results show that learned models are indeed more accurate than traditional methods, but they often suffer from high training and inference costs. Secondly, we explore whether these learned models are ready for dynamic environments (i.e., frequent data updates). We find that they cannot catch up with fast data updates and return large errors for different reasons. For less frequent updates, they can perform better but there is no clear winner among themselves. Thirdly, we take a deeper look into learned models and explore when they may go wrong. Our results show that the performance of learned methods can be greatly affected by the changes in correlation, skewness, or domain size. More importantly, their behaviors are much harder to interpret and often unpredictable. Based on these findings, we identify two promising research directions (control the cost of learned models and make learned models trustworthy) and suggest a number of research opportunities. We hope that our study can guide researchers and practitioners to work together to eventually push learned cardinality estimators into real database systems.

Medication-related osteonecrosis of the jaw (MRONJ) is a rare and severe adverse drug reaction (ADR) commonly seen in people taking drugs prescribed for metastatic cancer and osteoporosis. Prior studies only analyzed this ADR utilizing the public database Food and Drug Administration Adverse Event Reporting System (FAERS) by different stages (2010-2014, 2015-2021); a more comprehensive study is needed to analyze the MRONJ cases and the associated drugs from a longer time span. We conducted a retrospective pharmacovigilance analysis for all reported MRONJ cases between 2010 and 2023 in FAERS, using preferred terms and the primary suspect drug as searching conditions. Thus, this study aimed to analyze the MRONJ cases and the associated drugs more comprehensively. The reporting odds ratios (RORs) and 95% CIs were calculated for each queried drug. To distinguish the signal levels, we calculated the expected information component (EIC) and its 95% CI, using the Bayesian confidence propagation neural network (BCPNN) methods. We identified 22 846 MRONJ cases. A total of 15 drug classes including 30 suspect drugs showed different positive signal levels; among these drugs, 8, 5, and 17 had strong, medium, and weak intensity signals (+++, ++, and +), respectively. Drug classes involved bisphosphonate, RANKL inhibitor, radiotherapy drug, monoclonal antibody for cancer, corticosteroid, tyrosine kinase inhibitor, mammalian target of rapamycin inhibitor, aromatase inhibitor, cyclin-dependent kinase 4/6 inhibitor, immunomodulator, microtubule inhibitor, selective estrogen receptor modulator, sclerostin monoclonal antibody, estrogen receptor antagonist, and cytotoxic drug. Bisphosphonate, RANKL inhibitor, and radiotherapy drug exhibited higher risk than other classes with higher ROR or EIC 95% CI lower limit. Females had higher MRONJ incidence than males, and the mean age was 67.33 ± 11.71 yr. Compared with previous research, this study identified more drug classes and more novel medications with positive MRONJ signals that warrant further attention.