[The spring of artificial intelligence: AI vs. expert for internal medicine cases]

INTRODUCTION

The "Printemps de la Médecine Interne" are training days for Francophone internists. The clinical cases presented during these days are complex. This study aims to evaluate the diagnostic capabilities of non-specialized artificial intelligence (language models) ChatGPT-4 and Bard by confronting them with the puzzles of the "Printemps de la Médecine Interne".

METHOD

Clinical cases from the "Printemps de la Médecine Interne" 2021 and 2022 were submitted to two language models: ChatGPT-4 and Bard. In case of a wrong answer, a second attempt was offered. We then compared the responses of human internist experts to those of artificial intelligence.

RESULTS

Of the 12 clinical cases submitted, human internist experts diagnosed nine, ChatGPT-4 diagnosed three, and Bard diagnosed one. One of the cases solved by ChatGPT-4 was not solved by the internist expert. The artificial intelligence had a response time of a few seconds.

CONCLUSIONS

Currently, the diagnostic skills of ChatGPT-4 and Bard are inferior to those of human experts in solving complex clinical cases but are very promising. Recently made available to the general public, they already have impressive capabilities, questioning the role of the diagnostic physician. It would be advisable to adapt the rules or subjects of future "Printemps de la Médecine Interne" so that they are not solved by a public language model.

Pregnancy outcome in patients with a medical history of immunoglobulin A vasculitis: a case-control study

OBJECTIVE

Data on obstetric outcomes in patients with a history of immunoglobulin A vasculitis (IgA-V) are lacking. The aim of this study was to assess maternal, neonatal, and vasculitis outcomes during pregnancy.

METHOD

We conducted a French retrospective case-control study. Pregnancies of patients with a history of IgA-V (cases) were retrospectively studied and compared to pregnancies in women who developed IgA-V after their pregnancies and to pregnancies in healthy women (controls).

RESULTS

Twenty-six pregnancies in patients with a history of IgA-V were included and compared to 15 pregnancies in women who later developed IgA-V and 52 pregnancies in healthy women. Both gestational hypertension and pre-eclampsia were more frequent in the case group than in the other groups (23% vs 0% vs 0%, p < 0.01; 12% vs 7% vs 0%, p = 0.04). Hypertensive disorder of pregnancy occurred more frequently in patients with pre-existing kidney disease (78% vs 12%, p < 0.01). Caesarean section was more often performed in the case group than in the other groups (27% vs 0% vs 10%, p = 0.04). No foetal loss or maternal deaths occurred. There were no differences in delivery term or birth weight. No vasculitis flares were observed during pregnancy.

CONCLUSION

Women with a history of IgA-V appear to be at higher risk for gestational hypertension and pre-eclampsia, especially in cases with renal involvement; however, both mother and newborn outcomes appear to be favourable.

[Libman-Sacks endocarditis under apixaban in a patient with a high-risk profile venous antiphospholipid syndrome]

INTRODUCTION

Libman-Sacks endocarditis is a rare complication of antiphospholipid syndrome. Anti-vitamin K therapy is the standard treatment, although valvular replacement surgery may be required in some severe cases. In the latest EULAR recommendations, it is advised not to use direct oral anticoagulants in the management of antiphospholipid syndrome, especially of high-risk profile.

CASE REPORT

We present a case of a mitral Libman-Sacks endocarditis complicated with multiple strokes occurring in the setting of an antiphospholipid syndrome with triple positive antibody profile in a 63-year-old woman with multiple sclerosis. She was previously treated with apixaban for two years. Tinzaparin followed by prolonged warfarine treatment and two months of hydroxychloroquine resulted in valvular improvement.

CONCLUSION

To our knowledge, this is the first case of Libman-Sacks endocarditis occurring during apixaban therapy in a patient with antiphospholipid syndrome. This severe case highlights the inefficiency of direct oral anticoagulants to prevent thrombotic events in the antiphospholipid syndrome.

[A microcytic sideroblastic anemia successfully treated with B6 vitamin]

INTRODUCTION

Sideroblastic anemia is a rare cause of microcytic anemia, which is characterized by ring sideroblasts on bone marrow aspirate. This anemia can be congenital or acquired.

CASE REPORT

We report the case of an alcoholic 49-year-old man who presented with a severe microcytic sideroblastic anemia related to pyridoxine (B6 vitamin) deficiency. Acid folic deficiency was associated. The blood count normalized within one month after vitamin supplementation.

CONCLUSION

Pyridoxine deficiency must be sought in sideroblastic anemia in patients at risk.

Manic switches induced by antidepressants: an umbrella review comparing randomized controlled trials and observational studies

OBJECTIVE

We aimed to explore whether the prevalence of manic switch was underestimated in randomized controlled trials (RCTs) compared to observational studies (OSs).

METHOD

Meta-analyses and simple and systematic reviews were identified by two reviewers in a blinded, standardized manner. All relevant references were extracted to include RCTs and OSs that provided data about manic switch prevalence after antidepressant treatment for a major depressive episode. The primary outcome was manic switch prevalence in the different arms of each study. A meta-regression was conducted to quantify the impact of certain variables on manic switch prevalence.

RESULTS

A total of 57 papers (35 RCTs and 22 OSs) were included in the main analysis. RCTs underestimated the rate of manic switch [0.53 (0.32-0.87)]. Overestimated prevalence was related to imipraminics [1.85 (1.22-2.79)]; to serotonin-norepinephrine reuptake inhibitors [1.74 (1.06-2.86)]; and to other classes of drugs [1.58 (1.08-2.31)], compared to placebo treatment. The prevalence of manic switch was lower among adults than among children [0.2 (0.07-0.59)]; and higher [20.58 (8.41-50.31)] in case of bipolar disorder.

CONCLUSION

Our results highlight an underestimation of the rates of manic switch under antidepressants in RCTs compared to the rates observed in observational studies.