1. Lenze EJ, Mattar C, Zorumski CF, Stevens A, Schweiger J, Nicol GE et al (2020) Fluvoxamine vs placebo and clinical deterioration in outpatients with symptomatic COVID-19: a randomized clinical trial. JAMA 324:2292-2300.
  2. Reis G, dos Santos Moreira-Silva EA, Medeiros Silva DC, Thabane L, Cruz Milagres A et al (2022) Effect of early treatment with fluvoxamine on risk of emergeny care and hospitalizations among patients with COVID-19: the TOGETHER randomized platform trial. Lancet Glob Health 10: e42-51 https://doi.org/10.1016/S2214-109X(21)00448-4
  3. Lenze E. Fluvoxamine for early treatment of COVID-19: a fully-remote, randomized placebo controlled trial. ClinicalTrials.gov. Accessed September 26, 2022. https://clinicaltrials.gov/ct2/show/NCT04668950
  4. Seo H, Kim H, Bae S, Park S, Chung H, Sung HS et al (2022) Fluvoxamine treatment of patients with symptomatic COVID-19 in a community treatment center: a preliminary result of randomized controlled trial. Infect Chemother 54:102-113.
  5. Bramante CT, Juling JD, Tignanelli CJ, Buse JB, Liebovitz DM, Nicklas JM et al (2022) Randomized trial of metformin, ivermectin and fluvoxamine for Covid-19. N Engl J Med 387:599-610.
  6. McCarthy MW, Naggie S, Boulware DR, Lindsell CJ, Stewart TG, Felker GM et al (2022) Fluvoxamine for outpatient treatment of COVID-19: a decentralized, placebo-controlled, randomized platform clinical trial. medRxiv https://doi.org/10.1101/2022.10.17.22281178
  7. Sukhatme VP, Reiersen AM, Vayttaden SJ, Sukhatme VV (2021) Fluvoxiamine: a review of its mechanism of action and its role in COVID-19. Frontiers Pharmacol 12:652688. https://doi.org/10.3389/fphar.2021.652688
  8. Zheng W, Sun HL, Cai H, Zhang Q, Ng CH, Xiang YT (2022) Antidepressants for COVID-19: a systematic review. J Affective Disord 307:108-114.
  9. Mas M, Garcia-Vincente JA, Estrada-Gelnoch A, Perez-Mana C, Papaseit E, Torrens M et al (2022) Antidepressant drugs and COVID-19: a review of basic and clinical evidence. J Clin Med 11:4038. https://doi.org/10.3390/jcm11144038
  10. Strom BL (2021) Study design available for pharmacoepidemiological studies. In: Strom BL, Kimmel E, Hennessy S, eds. Textbook of pharmacoepidemiology, 3rd ed. Hoboken, NJ: Wiley-Blackwell, pp. 20-46.
  11. Vai B, Mazza MG, Delli Colli C, Foiselle M, Allen B, Benedetti F et al (2021) Mental disorders and risk of COVID-19-related mortality, hospitalization, and intensive care unit admission: a systematic review and meta-analysis. Lancet Psychiatry 8:797-812.
  12. Fond G, Nemani K, Etchecopar-Etchart D, Loundou A, Goff DC, Lee SW et al (2021) Association between mental health disorders and mortality among patients with COVID-19 in 7 countries: a systematic review and meta-analysis. JAMA Psychiatry 78: 1208-1217
  13. Griffith GJ, Morris TT, Tudball MJ, Herbert A, Mancano G, Pike L et al (2020) Collider bias undermines our understanding of COVID-19 disease risk and severity. Nat Commun11(1):5749 https://doi.org/10.1038/s41467-020-19478-2
  14. Ceban F, Nogo D, Carvalho IP, Lee Y, Nasri F, Xiong J et al (2021) Association between mood disorders and risk of COVID-19 infection, hospitalization and death. A systematic review and meta-analysis. JAMA Psychiatry 78:1079-1091.
  15. Mahdi M, Herman L, Rethelyi JM, Balint L (2022) Potential role of antidepressants fluoxetine and fluvoxamine in the treatment of COVID-19. Int J Mol Sci 23:3812. https://doi.org/10.3390/ijms23073812
  16. Oskotsky T, Marić I, Tang A, Oskotsky B, Wong RJ, Aghaeepour N et al (2021) Mortality risk among patients with COVID-19 prescribed selective serotonin reuptake inhibitor antidepressants. JAMA Network Open 4(11):e2133090. https://doi.org/10.1001/jamanetworkopen.2021.33090
  17. Ho DE, Imai K, King G, Stuart EA (2011) MatchIt : Nonparametric Preprocessing for Parametric Causal Inference. J Stat Softw42(8):1-28. https://doi.org/10.18637/jss.v042.i08
  18. R Core Team (2020) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.
  19. Trkulja V (2022) Fluvoxamine for COVID-19 outpatients: for the time being, we might prefer to curb our optimism. Br J Clin Pharmacol 88:4654-4656.
  20. Goodrich B, Gabry J, Ali I, Brilleman S (2022) rstanarm: Bayesian applied regression modeling via Stan. R package version 2.21.3 https://mc-stan.org/rstanarm/
  21. Schwarzer G, Carpenter JR, Rucker G (2015) Meta-analysis with R. Springer International Publishing Switzerland, 2015, pp.187-216.
  22. Beliveau A, Boyne DJ, Slater J, Brenner D, Arora P (2019) BUGSnet: an R package to facilitate the conduct and reporting of Bayesian network meta-analyses. Med Res Methodology 19:196. https://doi.org/10.1186/s12874-019-0829-2
  23. van Valkenhoef G, Lu G, de Brock B, Hillege H, Ades AE, Welton J (2012) Automating network meta-analysis. Res Synth Methods 3:285-299.
  24. Schneeweiss S (2006) Sensitivity analysis and external adjustment for unmeasured confounders in epidemiologic database studies of therapeutics. Pharmacoepidemiol Drug Saf 15:291-303.
  25. Heine D (2021). The episensr package: basic sensitivity analysis of epidemiological results. doi:10.5281/zenodo.4554553, R package version 1.1.0 https://dhaine.github.io/episensr/
  26. Kuran BT, Greer TL, Trivedi MH (2010) Strategies to enhance the therapeutic efficacy of antidepressants: targeting residual symptoms. Expert Rev Neurother 9:975-984.
  27. Ros S, Aguera L, de la Gandara J, Rojo JE, de Pedro JM (2005) Potentiation strategies for treatment-resistant depression. Acta Psyhciatr Scand Suppl 35:14-24.
  28. Ballenger JC (1999) Clinical guidelines for establishing remission in patients with depression and anxiety. J Clin Psyciatry 60(Suppl. 22):29-34.
  29. Hadley E, Rhea S, Jones K, Li L, Sotner M, Bobashev G (2022). Enhancing the prediction of hospitalization from a COVID-19 agent-based model: a Bayesian method for model parameter estimation. PloS ONE 17(3):e0264704. https://doi.org/10.1371/journal.pone.0264704
  30. Herrera-Esposito D, de los Campos G (2022) Age-specific rate of severe and critical SARS-CoV-2 infections estimated with multi-country seroprevalence studies. BMC Infect Dis 22:311. https://doi.org/10.1186/s12879-022-07262-0
  31. Croatia: Coronavirus Pandemic Country Profile - Our World in Data (accessed September 20, 2022).

Table 1. Subsets of people diagnosed with COVID-19 (up to August 15, 2021) in respect to exposure to fluvoxamine around the time of COVID-19 diagnosis.

Study 1

  • Cohort A. People suffering difficulties requiring antidepressants/anxyolytics and prescribed/ exposed to fluvoxamine – (i) at least one International Classification of Disease (ICD-10) F or G30/31.1 code entry (conditions in which fluvoxamine or other antidepressants/anxiolytics might have been the main or one of the required treatments) at any time between January 1, 2019 up to 7 days after the date of the index COVID-19 diagnosis; (ii) at least one prescription for fluvoxamine issued in the period betrween 90 days prior to- and 7 days after the index COVID-19 diagnosis; (iii) could have been prescribed other psychiatric treatments including other antidepressants/ anxiolytics between January 1, 2019 and the date of the index COVID-19 diagnosis.

  • Cohort B. People suffering difficulties requiring antidepressants/anxyolytics not prescribed/ exposed to fluvoxamine – (i) at least one ICD-10 F or G30/G31.1 code entry (conditions in which antidepressants/anxiolytics might have been the main or one of the required treatments) at any time between January 1 2019 up to 7 days after the date of the index COVID-19 diagnosis; (ii) no prescription for fluvoxamine issued in the period between 6 months prior to- and 21 days after the index COVID-19 diagnosis; (iii) could have been prescribed other psychiatric treatments including other antidepressants/ anxiolytics between January 1, 2019 and the date of the index COVID-19 diagnosis.

  • Cohort C. People free of psychiatric difficulties and not prescribed/exposed to fluvoxamine or to any other pharmacological psychiatric treatment – (i) no ICD-10 F code entries at any time between January 1 2019 and 21 days after the index COVID-19 diagnosis, and (ii) no prescriptions for fluvoxamine or any of the other drugs falling into the Anatomical Therapeutic Chemical codes N05, N06 or N07B in the period between 6 months prior to- and 21 days after the index COVID-19 diagnosis.

Study 2

  • Cohort Fluvoxamine. The same as Cohort A in Study 1, except that “other treatments” exclude paroxetine: no prescriptions issued between 6 months prior to- and 21 days after the index COVID-19 diagnosis.

  • Cohort Paroxetine. People suffering difficulties requiring antidepressants/anxyolytics and prescribed/exposed to paroxetine – (i) at least one ICD-10 F or G30/31.1 code entry (conditions in which paroxetine or other antidepressants/anxiolytics might have been the main or one of the required treatments) at any time between January 1, 2019 up to 7 days after the date of the index COVID-19 diagnosis; (ii) at least one prescription for paroxetine issued in the period betrween 90 days prior to- and 7 days after the index COVID-19 diagnosis; (iii) could have been prescribed other psychiatric treatments including other antidepressants/ anxiolytics between January 1, 2019 and the date of the index COVID-19 diagnosis, except for fluvoxamine: no prescription issued between 6 months prior to- and 21 days after the index COVID-19 diagnosis.

Table 2. Covariates used for exact matching between patient subsets (Cohorts based on burden of psychiatric conditions and exposure to fluvoxamine/paroxetine).

Matching variables used for all comparisons

Age

As 5-year bins between 16 and 111 years

Sex

Male or female

Vaccination status

Not vaccinated; received a single-dose vaccine (i) <14 days before COVID-19 diagnosis; ii) 14-90 days before or (iii) >90 days before COVID-19 diagnosis; received the 1st dose of a two-dose vaccine (i) <14 days before; (ii) 14-90 days before or (iii) >90 days before COVID-19 diagnosis; received the 2nd dose of a two-dose vaccine (i) <14 days before; (ii) 14-90 days before or (iii) >90 days before COVID-19 diagnosis

Calendar period

Up to January 9 2020 (including) – still no vaccination, Alpha strain(s) prevailing; January 10 – July 15 2021 – Alpha strain(s) prevailing, mass vaccination in progress; after July 15 2021 – Delta strain starts to prevail, mass vaccination in progress.

Comorbidities

Charlson comorbidity index (CCI) subseted at 4 levels: 0, 1-2, 3-4 and ≥5, and also individual comorbidities: atrial fibrillation/undulation, autoimmune disease, malignant disease (cancer), congestive heart failure, chronic obstructive pulmonary disease, history of ischemic heart disease or a cerebrovascular disease, renal disease (in addition to codes in CCI: chronic kidney disease, N18; and dependence on renal dialysis, Z99.2), diabetes without complications, diabetes with complications and dementia (same ICD-10 codes as for the calculation of CCI).

Pharmacological treatments

Inhibitors of the renin-angiotensine-aldosterone system (RAAS) (includes any of the following: angiotensine converting enzyme inhibitors, angiotensine receptor antagonists and mineralocorticoid receptor antagonists); diuretics (any).

Matching variables additionally used in the comparison between patients burdened with psychiatric difficulties that may require antidepressant/anxyolytic treatment. In Study 1 this refers to Cohorts A and B (Cohort C by definition is free of such conditions). In Study 2, it refers to both cohorts.

Mood disorders

Mood (affective) disorders (F30-F39)

Nonpsychotic mood disorders

Anxiety, dissociative, stress-related, somatoform and other nonpsychotic mood disorders (F40-F48)

Substance use

Mental and behavioral disorders due to psychoactive substance use (F10-F19)

Non-mood psychotic disorders

Schizophrenia, schizotypal, delusional and other non-mood psychotic disorders (F20-F29)

Cumulatively: F50-F59, F60-F69, F70-F79, F80-F89, F90-F98, F04-F09

Behavioral syndromes associated with physiological disturbances and physical factors; Disorders of adult personality and behavior; Intellectual disabilities; Pervasive and specific developmental disorders; Behavioral and emotional disorders with onset in childhood and adolescence; mental disorders due to know physiological condition (F04,05,06, 07, 09)

Table 3 Subject characteristics across subsets based on burden of psychiatric conditions and exposure to fluvoxamine/paroxetine (see Table 1) – before matching [n (%), except age]. Shown are all covariates used for matching (see Table 2). “Vaccination status” refers to number of doses received/number needed for full vaccination and time elapsed since the last vaccine dose.

Study 1

Study 2

Cohort A

Cohort B

Cohort C

Fluvoxamine

Paroxetine

N

1016

95984

275804

994

1796

Age (years)1 [meanSD (range)]

5518 (16-99)

5817 (16-105)

4316 (16-96)

5517(16-98)

5816(16-95)

Male

441 (43.4)

33604 (35.0)

144548 (52.4)

431 (43.4)

564 (31.4)

Vaccination status

1/1, >90 days

0 (0.0)

1 (<0.1)

1 (<0.1)

0 (0.0)

0 (0.0)

1/1, <14 days

0 (0.0)

13 (<0.1)

38 (0.1)

0 (0.0)

0 (0.0)

1/1, 14-90 days

0 (0.0)

8 (<0.1)

43 (0.2)

0 (0.0)

0 (0.0)

Not vaccinated

958 (94.3)

90719 (94.5)

269392 (97.7)

936 (94.2)

1697 (94.5)

1/2, >90 days

0 (0.0)

28 (0.0)

16 (0.0)

0 (0.0)

0 (0.0)

1/2, <14 days

23 (2.3)

1724 (1.8)

2667 (1.0)

23 (2.3)

27 (1.5)

1/2, 14-90 days

27 (2.7)

2431 (2.5)

2645 (1.0)

27 (2.7)

51 (2.8)

2/2, >90 days

1 (0.1)

237 (0.2)

201 (0.1)

1 (0.1)

5 (0.3)

2/2, <14 days

0 (0.0)

298 (0.3)

205 (0.1)

0 (0.0)

4 (0.2)

2/2, 14-90 days

7 (0.7)

525 (0.5)

596 (0.2)

7 (0.7)

12 (0.7)

Calendar period

Up to January 9, 2021

591 (58.2)

55623 (58.0)

163836 (59.4)

574 (57.7)

1061 (59.1)

January 10 to July 15, 2021

408 (40.2)

39112 (40.7)

106379 (38.6)

403 (40.5)

716 (39.9)

After July 15, 2021

17 (1.7)

1249 (1.3)

5589 (2.0)

17 (1.7)

19 (1.1)

Weighted CCI

0

609 (59.9)

55122 (57.4)

230007 (83.4)

598 (60.2)

981 (54.6)

1-2

293 (28.8)

31229 (32.5)

40370 (14.6)

285 (28.7)

648 (36.1)

3-4

86 (8.5)

7375 (7.7)

4380 (1.6)

85 (8.6)

132 (7.3)

5

28 (2.8)

2258 (2.4)

1047 (0.4)

26 (2.6)

35 (1.9)

Additional individual conditions

Atrial fibrillation/undulation

61 (6.0)

5850 (6.1)

4017 (1.5)

60 (6.0)

84 (4.7)

Autoimmune disease

112 (11.0)

13295 (13.9)

14281 (5.2)

110 (11.1)

268 (14.9)

Cancer

58 (5.7)

7569 (7.9)

7109 (2.6)

55 (5.5)

128 (7.1)

Congestive heart failure

30 (3.0)

3907 (4.1)

2149 (0.8)

30 (3.0)

66 (3.7)

COPD

112 (11.0)

12808 (13.3)

17171 (6.2)

109 (11.0)

269 (15.0)

IHD or CVD

149 (14.7)

12803 (13.3)

8240 (3.0)

146 (14.7)

241 (13.4)

Renal disease2

23 (2.3)

2067 (2.2)

1334 (0.5)

23 (2.3)

31 (1.7)

Diabetes with complications

19 (1.9)

1611 (1.7)

1131 (0.4)

18 (1.8)

34 (1.9)

Diabetes w/o complications

165 (16.2)

15155 (15.8)

15706 (5.7)

159 (16.0)

315 (17.5)

Dementia

37 (3.6)

2356 (2.5)

2482 (0.9)

37 (3.7)

56 (3.1)

Immunocompromised

14 (1.4)

1559 (1.6)

1375 (0.5)

14 (1.4)

23 (1.3)

Using RAAS inhibitors

293 (28.8)

30443 (31.7)

30386 (11.0)

284 (28.6)

644 (35.9)

Using diuretics

125 (12.3)

14193 (14.8)

9719 (3.5)

121 (12.2)

281 (15.6)

F10-F19

28 (2.8)

2650 (2.8)

---

27 (2.7)

44 (2.4)

F20-F29

119 (11.7)

5306 (5.5)

---

117 (11.8)

112 (6.2)

F30-F39

446 (43.9)

20856 (21.7)

---

434 (43.7)

971 (54.1)

F40-F48

692 (68.1)

71242 (74.2)

---

676 (68.0)

1199 (66.8)

Any of F04-F09, F50-F59, F60-F69, F70-F79, F80-F89, F90-F98

229 (22.5)

19431 (20.2)

---

218 (21.9)

307 (17.1)

Outcomes

COVID-related hospitalization

34 (3.35)

3128 (3.25)

2590 (0.94)

32 (3.22)

65 (3.62)

All-cause 30-day hospitalization

127 (12.5)

11266 (11.7)

14297 (5.18)

125 (12.6)

206 (11.5)

COVID-related mortality (composite)

38 (3.74)

4261 (4.44)

2898 (1.05)

37 (3.72)

80 (4.44)

1 For clarity, age is summarized. In the matching process, it was binned (see Table 2).

CCI – Charlson comorbidity index; COPD – chronic obstructive pulmonary disease; CVD – cerebrovascular disease; IHD – ischemic heart disease; RAAS – renin angiotensin aldosteron system

2 In addition to codes in CCI: chronic kidney disease, N18; and dependence on renal dialysis, Z99.2

Table 4 Analysis of sensitivity to unmeasured confounding. Estimates of the effect of exposure to fluvoxamine (Cohort A in Study 1, Cohort Fluvoxamine in Study 2) vs. non-exposure (control condition; Cohort B in Study 1, Cohort Paroxetine in Study 2) generated in the Bayesian analysis with a moderately informative skeptical prior were corrected for (hypothetical) unmeasured confounding that diminished the (presumed) risk-reducing effect of fluvoxamine: a large imbalance (1:40) between fluvoxamine-exposed and control subjects was hypothesized in prevalence of an “other co-treatment” (e.g., other antidepressant/anxiolytic) with a high efficacy (30% risk reduction) against COVID-19 disease progression. Shown are actually observed and bias-corrected estimates of relative risks (RR) for all outcomes in Study 1 and Study 2.

Observed RR (95%CrI)

Bias-corrected RR (95%CrI)

Study 1

Cohort A vs. Cohort B

COVID-19-related hospitalization

1.15 (0.66-2.11)

1.02 (0.58-1.86)

All-cause 30-day hospitalization

1.76 (1.39-2.25)

1.55 (1.23-1.99)

COVID-19-related mortality

0.93 (0.53-1.76)

0.82 (0.47-1.55)

Study 2

Fluvoxamine vs. Paroxetine

COVID-19-related hospitalization

1.21 (0.60-2.36)

1.07 (0.52-2.08)

All-cause 30-day hospitalization

1.13 (0.73-1.73)

1.00 (0.64-1.53)

COVID-19-related mortality

0.91 (0.46-1.72)

0.80 (0.41-1.52)