We conducted a multi-center retrospective cohort study of adults (aged ≥ 18 years) who were diagnosed with nocardiosis between 2007 and 2022.

The cohort comprised individuals from four tertiary medical centers in Israel: Rabin Medical Center– Beilinson Hospital, Sheba Medical Center, Rambam Healthcare Campus, and Soroka Medical Center. Characteristics of participating centers and periods of patients’ identification are provided in supplementary Table S1. We reviewed the microbiology laboratory records of participating centers and retrieved the medical charts of all individuals from whom Nocardia was isolated up to December 31st, 2022, allowing at least one-year follow-up for all included individuals.

We defined nocardiosis as Nocardia species isolation from an individual who received a formal diagnosis of nocardiosis by an infectious diseases specialist and was treated with at least one agent directed against Nocardia species, either empirically or as guided by antibiotic susceptibility testing. Nocardia colonization was defined as Nocardia isolation from an individual who had no relevant symptoms, was not diagnosed with nocardiosis, and did not receive treatment directed at Nocardia [8]. We included all individuals with nocardiosis and excluded all cases of Nocardia colonization. Individuals with nocardiosis for whom data on treatment duration were unavailable or could not be ascertained were also excluded.

Our primary exposure variable was the total antibiotic treatment duration, defined as the period in which treatment directed against nocardiosis was given, regardless of antibiotic type, number of agents, and route of administration, as these could change throughout treatment (e.g., switching from combination to monotherapy). Secondary prophylaxis with oral low-dose SXT (i.e., 800/160 mg once daily or less frequently, following a treatment period at therapeutic doses or other agents) was not counted as treatment.

Based on the previously suggested treatment algorithm [2], we classified the cohort into three treatment duration groups: short (≤ 90 days), intermediate (91–180 days), and prolonged (> 180 days). Intravenous treatment duration (defined as the period in which at least one agent was administered intravenously) and combination therapy duration (i.e., the period in which ≥ 2 agents were administered) were also calculated.

We collected data on patients’ baseline characteristics, including demographics, comorbidities, and immune status [see below]. Information on nocardiosis included the causative Nocardia species, the infection syndrome [see below], and presentation (need for oxygen and markers of inflammation). Data on surgical interventions for either diagnostic workup or therapeutic purposes (debridement, drainage of an abscess, heart valve replacement, etc.) were also collected.

We classified patients into three immune status groups, based on their underlying comorbidities and pharmacotherapy (i.e., use of immunosuppressive or immunomodulatory agents), according to the level of immune suppression: substantial immune suppression (SOTRs, hematopoietic stem cell transplant recipients [HCTs], individuals with acquired immunodeficiency syndrome [AIDS], and those on high-dose corticosteroid therapy [≥ 20 mg prednisone equivalent per day] for at least 21 days prior to nocardiosis); mild to moderate immune suppression (individuals with active malignancy, primary immune deficiency, autoimmune diseases, usage of immunomodulatory agents and those on medium dose corticosteroid therapy [≥ 5 and < 20 mg prednisone equivalent per day] for at least 21 days) prior to nocardiosis; and apparently immunocompetence (all others).

Nocardiosis was classified into one of five syndromes: solitary lymphocutaneous nocardiosis (defined as an infection involving skin and soft tissue only); Solitary pulmonary nocardiosis (defined as an infection involving only the lungs’ parenchyma, with or without empyema); disseminated nocardiosis without central nervous system (CNS) involvement (defined as involvement of ≥ 2 non-contiguous organs or Nocardia bacteremia), disseminated nocardiosis with CNS involvement (defined as any infection involving the CNS), and Nocardia osteomyelitis.

Our primary outcome was 1-year all-cause mortality.

Secondary outcomes included nocardiosis relapse and antibiotic-related adverse events within one year.

Nocardiosis relapse was defined as the recurrence of signs and symptoms, potentially following resolution period, with repeated microbiological evidence of the same Nocardia species.

Antibiotic-related adverse events were treated as a composite outcome, including at least one of the following: renal toxicity, bone marrow toxicity, Clostridioides Difficile infection (CDI), and allergic reactions related to the regimen used against nocardiosis (definitions for each of the four adverse events are provided in the supplementary material (supplementary Table S2).

Information on the study variables was manually extracted from patients’ medical charts. We included all hospital visits (either admissions or outpatient clinics) during the year following the diagnosis and recorded all antibiotic therapy, dates of regimen alternations (in either agent, dosage, or route of administration), dates of treatment secession, and microbiology (e.g., repeated cultures).

Data collected through different sources (hospital and community records) were cross-checked against each other to guarantee their consistency and reliability.

Sputum, bronchoalveolar lavage, blood, and tissue samples from infected individuals were incubated on blood and chocolate agar plates for at least 2–4 days. When Nocardia was suspected, samples were seeded on additional plates (such as Tayer–Martin agar or Buffered Charcoal Yeast Extract) and incubated for longer durations.

Nocardia identification was based on macroscopic and microscopic assessment combined with biochemical markers early during the study years and matrix-assisted laser desorption/ionization time-of-flight mass spectrometer (MALDI-TOF MS) or sequencing when it became available. Whenever performed, antimicrobial susceptibility testing was performed using either broth microdilution or E-testing (supplementary Table S3).

Treatment duration was cross-tabulated with immune status, nocardiosis syndromes, Nocardia species, and other demographic and clinical characteristics. Treatment duration groups were compared using Chi square or Fisher’s exact tests for categorical variables and one-way ANOVA or the Mann-Whitney-U test for continuous variables. Kaplan–Meier survival curves were plotted and assessed for the primary outcome using the Log-rank test.

Considering the immense heterogeneity in clinical syndromes and host-pathogen relationships, making nocardiosis cases non-comparable by no means, we decided to renounce adjustments, recognizing that clinical judgment has led clinicians to shorten treatment durations in appropriate cases. Accordingly, multivariable analysis was not conducted.

To validate our findings, we implemented several sensitivity analyses:

Considering that we excluded individuals who died during the 90 days from nocardiosis diagnosis while on treatment, we reanalyzed the data following an exclusion of all individuals who died while on therapy, at any time point (i.e., even beyond 90 days).

Additionally, we reanalyzed our data following reclassification of the treatment duration groups, once by including those who were treated for 91–120 days in the short treatment group, and once by combining the short and intermediate groups (i.e., short-intermediate vs. prolonged treatment duration, using a single cutoff of 180 days).

For all analyses, p < 0.05 was considered statistically significant.

We analyzed data using IBM SPSS version 28 (Armonk, New York, USA) and GraphPad Prism 8.0.1.

The study was approved by the Institutional Review Boards (IRB) of each participating center (RMC-0685-21; SMC-9959-22; RMB-D-0039-23; SOR-0046-23). The IRBs granted an exemption from obtaining informed consent due to the study’s retrospective design, and we followed the STROBE guidelines for observational studies [9].