Presence and detectability of the second victim phenomenon in German-speaking medical professionals involved in dysphagia and respiratory therapy: cross-sectional survey research (SeViD-V) Authors: Stefan Bushuven, Tobias P. Taubner, Milena Trifunovic-Koenig, Paul Diesener, Bianka Gerber, Christina Rohlfes, Silke Kuske, Hannah Roesner, Stefanie Bushuven, Reinhard Strametz Published: 2026 ABSTRACT BackgroundSecond victim effects refer to experiences of severe adverse events and errors in healthcare that have a negative impact on healthcare professionals. Speech and language pathologists and respiratory therapists are involved in treating patients with dysphagia in intensive care and stroke units and are thus exposed to possible severe incidents related to choking, aspiration, resuscitation, palliative care, and death. Evidence of the detectability of the second victim phenomenon in this professional group remains limited—a gap addressed in this study.MethodsUsing convenience sampling, an online questionnaire was distributed among medical professionals involved in dysphagia therapy in German-speaking countries. Questionnaire components included screening tools and psychometric instruments. Data analysis involved descriptive statistics, regression analyses, and content analysis of free-text entries.ResultsNinety-one of 127 speech and language pathologists and respiratory therapists (71,65%) reported having experienced a second victim phenomenon at least once in their lives. Twelve participants reported having experienced it for more than one year. Nine participants who denied having experienced a second victim effect initially described symptoms in line with the consensus-based definition later in the free-text entries. Qualitative analysis revealed new coping strategies, including support from first victims and educating others.ConclusionOverall, second victim effects were detected among speech and language pathologists and respiratory therapists in German-speaking countries. The high prevalence and incidence rates found in this study are comparable to those reported in other studies on second victims in Germany. A newly identified dimension is the role of “silent victims,” including those who do not remember or deny being affected indicating the high demand for institutional prevention, support programs, and refinement of tools for optimized detectability. There is a need to improve the accuracy of second victim screening tools to avoid false-negatives of identifying second victims.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12913-026-14648-2. Dysphagia is a common disorder associated with several conditions and treatments, including stroke, intracranial bleeding, neurodegenerative diseases, traumatic brain injury, malignant tumor of the head and neck, neck surgery, radiation therapy, pulmonary diseases such as chronic obstructive pulmonary disease (COPD) with respiration-swallowing disorder, metabolic and neuropathic diseases, genetic inheritance diseases, and degenerative disease [1–4]. Dysphagia affects approximately 4%−6% of the US population, with prevalence rates increasing from 2012 to 2022 [2, 5]. Therapists, nurses, and doctors often face the dilemma of protecting airway safety on the one hand and patient well-being and social participation on the other [6]. In addition, life-threatening complications of chronic dysphagia (e.g. choking with asphyxia, pneumogenic sepsis) and dysphagia-related conditions (e.g., airway management in tracheostomy) may occur without warning [7, 8], leading to decisions prone to bias and error. For patients at risk of choking or aspiration, such as those with stroke, degenerative cerebral disease, critical illnesses and patients with tracheostomy, health care providers’ decisions on oral intake and airway management may be susceptible to errors or complications implicating the need for ongoing precise diagnostic and reevaluation [9]. Calibration between the over- and underestimation of pathology and its changes over time can be challenging. Erroneous decisions might lead to negative consequences in different ways, resulting in negative experiences for healthcare providers, such as regret, shame, feeling responsible for the errors, and guilt. Healthcare providers’ feelings of being impacted by an adverse event or a critical situation in patient care is known as the “second victim phenomenon” (SVP) [10] or, known in the context of resuscitation and subsequent progress of impairment as the “wounded healer”. In 2000, Albert Wu introduced the term “second victim” [11], with reference to a physician who has committed an error and subsequently experiences emotional suffering. After several extensions of the term and descriptions of multiple phases that can lead to dysfunctional behavior, job resignation, or even suicidal tendencies [12, 13], the European Researchers Network on Second Victims (ERNST) introduced a new definition in 2022 [10]: “Any healthcare worker, directly or indirectly involved in an unanticipated adverse patient event, unintentional healthcare error, or patient injury, and becomes victimized in the sense that also the worker is negatively impacted.” The inability to overcome feelings related to adverse events results in dysfunctional behavior, work-productivity-loss(absenteeism, presenteeism), and defensive medicine, further impairing psycho-emotional and patient safety [12, 14]. Higher prevalence and more severe adverse events may trigger reactions such as burnout, compassion fatigue, post-traumatic stress disorder, major depression, and even suicide [15]. Additional, moral distress and moral injury (deep violations of moral beliefs) may also play a role in second victim experience as it may mediate or modulate symptom load [16]. In Germany and Austria, several second victim studies have revealed self-reported lifetime prevalence rates of approximately 60-80% in participating young physicians [17] and nurses [18, 19], 53% in emergency physicians [20], 58% in family caregivers [21], 89% in pediatricians [22], 94% in midwives [23] and a high symptom load in general practitioners [16]. Further, neuroticism (emotional lability and susceptibility to negativity) is a possible risk factor for reporting victimization [19, 20]. Additionally, self-management of second victim effects may be influenced by the overconfidence effect, a cognitive bias suggesting that own capabilities are superior to others [24]. As such, factors such as overconfidence (“I can care for myself better than others can for themselves”), assumptions of under-reporting being a second victim, social desirability (“I have to be strong”), and label-avoidance (“I am not a victim”) may occur. These tendencies raise the question of whether asking a person if they feel affected by a medical error or adverse event or not is sufficient to detect the phenomenon. For measurement, a questionnaire called “Second Victims im Deutschsprachigen Raum/Second Victims in German-Speaking Countries (SeViD)” has been used. It is a content-validated German-language questionnaire to assess the prevalence, symptoms, risk factors, and support strategies related to the second victim phenomenon among healthcare professionals. Prior studies [17, 19] employed this tool using gated follow-up questions (only individuals affirming that they were affected were asked about related symptoms). Notably, the results of the SeViD IX study in general practitioners without gated questions showed that the identification of second victims may be underreported [16]. Germany has more than 31,000 speech and language therapists [25] (with a three-year primary education or bachelor’s degree) and approximately 1,500 respiratory therapists (with 18- to 24-month curricular education based on nursing education) [26, 27]. Both are working with dysphagic patients with speech and language therapists mainly but not exclusively dealing with dysphagia “above the vocal cords” (upper airway function and clearing, swallowing, nutrition, language) and respiratory therapists “from below the vocal cords” (lung function, clearing of lower airways, respirator therapy). Both are involved in patients with combined or overlapping conditions and assisting devices like tracheostomy. However, the number of these professionals actively and regularly involved in dysphagia treatment and concomitant conditions like tracheostomy remains unclear, as the treatment of dysphagia constitutes only one aspect of their work. Anticipating the high risks of second victim effects associated with dysphagia and respiratory therapy, there is a demand for investigate second victim effects in both domains. A previous study by Allender reported second victim effects in 171 respiratory therapists in the USA, using the “Second Victim Experience and Support Tool−Revised (SVEST-R) questionnaire” [28]. Further, Saragosa et al. (2024) identified moral distress in semi-structured interviews with 27 respiratory therapists in the USA [29]. Nonetheless, currently, a knowledge gap exists concerning professionals involved in dysphagia recognition and therapy. These are mainly but not exclusively, speech and language pathologists, respiratory therapists as well as specialized physicians and others involved; specifically, the presence and detectability of the SVP among these professionals treating dysphagia and respiratory diseases in Germany, Austria, and Switzerland remains unclear. Objectives This study explored the presence of second victim effects among healthcare providers treating patients with dysphagia in German-speaking countries. By combining quantitative and qualitative methods, it explored the detectability of the effects to investigate persons not identified as second victims, but suffering from related effects. Furthermore, we sought to explore additional effects concerning dysphagia therapy to generate hypotheses for future research on second victim effects in the field of airway safety and dysphagia therapy. Additional factors included personality traits, especially neuroticism being predictor in preceding studies [19, 20]. Further, we sought to implement the aspects of moral injury, as this was a factor in a prior study as well [16]. For secondary analyses, we performed multiple binary logistic regressions, with the outcome variable being having experienced an second victim effect. The model included the means of the five BFI-10 subscales, age, and sex as predictors, with continuous parameters tested for linearity.” Methods Study design We conducted an online cross-sectional anonymous survey of medical staff involved in the diagnosis and treatment of swallowing disorders in German-speaking countries, predominantly speech- and language pathologists and respiratory therapists. The survey included 34 questions pretested by five members of the authors’ group, consisting of three physicians (two with education and training in fiberoptic endoscopic evaluation of swallowing (FEES)), a speech and language therapist with a master’s degree in nutrition, and a nursing specialist. The nurse and therapists are both specialized in respiratory therapy and the management of swallowing disorders. Before recruitment, we consulted the ethics committee of the Physicians Association of Baden-Württemberg: no formal approval was required under national regulations. All participants were informed of the study objectives and agreed to participate. Data safety was ensured by following the General Data Protection Regulation (GDPR) and the German Data Protection Act (DGSVO). Setting The study was conducted from October 2023 to December 2024. The survey was administered in officially listed neurorehabilitation clinics and dysphagia networks in three German-speaking countries (Germany, Austria, and Switzerland). We employed an open convenience sampling method to collect responses. Participants were recruited via email to the heads of the departments of registered hospitals and institutions or via organizational contact forms and network newsletters. The heads of departments and network administrators were requested to forward the email to the persons involved. No additional advertisements or incentives were provided. Participants Speech language therapists, “patho-linguists”, intensive care nurses, respiratory therapists, and other healthcare professionals (physicians and occupational therapists) involved in dysphagia and respiratory therapy were eligible participants. We only included persons aged 18 years or older. The exclusion criteria were as follows: 1) participants who refused to share their data for scientific research; 2) those who answered only the demographic questions; 3) participants who omitted the question about having experienced a second victim effect; 4) “web-side visitors” (without interaction with the survey platform), and others who provided no meaningful information. Measurement The survey included 31 nonrandomized questions used in prior studies on second victims. To reduce the response burden, we used a modified SeViD questionnaire [17] to detect subjective second victim experiences (e.g., symptoms). We also used the German Version of the “SVEST-R (GSVEST-R)” [30, 31] to assess objective second victim experiences (e.g., psychological and physical distress, support factors, self-efficacy, resilience), the “Big Five Inventory–10 (BFI-10)” to measure personality traits (neuroticism, extraversion, openness to new experience, agreeableness, and conscientiousness [32], and the German Version of the “Moral Injury Symptom and Support Scale for Health Professionals (G-MISS-HP)” [33, 34] to measure moral injury (deep violations of moral beliefs).The term second victim was introduced and briefly defined in Question 2 (knowledge of the effect) and assessed in Questions 12 (being affected or not) and 13 (symptoms after critical events). Professional experiences with dysphagia were assessed using Questions 16 and 17. We displayed Items blockwise on 17 pages to limit the screen burden (see the Supplementary Materials). The platform did not provide completeness checks, but the participants were allowed to scroll back and forward through the screens. Follow-up questions were not gated. Thus, all participants (including those initially denying being affected) were asked questions regarding the effect duration, help consultation, recovery, symptoms falling under the definition of a second victim effect, and situations to which the experiences were related. All questions could be omitted when participants did not want to answer. For the questions on symptoms and incidents, we included “there was no event” as a possible answer. IP addresses were anonymized, but cookies were used to allow users to pause and restart the questionnaire and limit multiple accesses. The questionnaire is included in the Supplementary Materials. We report our study following the CHecklist for Reporting Results of Internet E-Surveys (CHERRIES) [35] and STrengthening the Reporting of OBservational Studies in Epidemiology (STROBE) [36] guidelines. Quantitative statistical methods We collected data via http://www.Umfrageonline.com (Enuvo GmbH, Zuerich, Switzerland) in conformance with the “European GDPR”. Data extraction, visualization, and clearing and descriptive analyses were performed using Microsoft Excel (Microsoft Corporation, Redmond, Washington, USA). Data management, particularly the recording of items, was cross-validated by a physician using artificial intelligence (ChatGPT 5.0). For analytical statistics, we used “R” (CRAN, 4.4.2, 2024–10-31) and R Studio (Posit 2024.12.0) with the packages tidyr, dplyr, psych, summarytools, skimr, lmtest, car, ggplot2, epiR, and DescTools. To verify the results and for regression analyses, we used JASP 0.95.1 (University of Amsterdam, the Netherlands) and SPSS 29.0 (IBM, Armonk, New York, USA). We described numerical data with means (“mean”) and standard deviations (“SD”), ordinal data with medians (“md”) and interquartile ranges (IQRs), and categorical data using numbers (“n”) and percentages rounded to one digit (%). To detect the phenomenon under study, we used descriptive statistics to address the question of being affected or not (“detection”). Further, we evaluated symptoms that fulfill the definition of an SVP, and analyzed free-text entries (“detectability”), particularly from responses of individuals who denied experiencing a second victim effect. For secondary analyses, we performed multiple binary logistic regressions, with the outcome variable being having experienced an SVP. The model included the means of the five BFI-10 subscales, age, and sex as predictors, with continuous parameters tested for linearity. For the model, we condensed the outcome variable of having experienced an SVP into two categories (0 = no SVP, 1 = SVP once or multiple times). For regression analysis, we excluded data from one non-binary participant and two participants with missing values (i.e., who did not provide information about their sex) leading to model instability. We assessed the explained variance using Nagelkerkes’ pseudo-R2 and excluded multicollinearity using the variance inflation factor (VIF) and tolerance values. Multicollinearity was assumed for tolerance < 0.1 and VIF > 5. The significance level was set at α < 0.05. We calculated the bias-corrected and accelerated bootstrapped confidence intervals based on 5,000 samples. Missing values were excluded listwise. As this was an explorative study, we did not adjust for multiplicity. Qualitative methods We analyzed free-text entries using content analysis [37], following a constructivist epistemological approach. This approach enabled a deeper evaluation, detection of interactions seldom observed, detection of the phenomena in persons who reported not being affected, and generation of new hypotheses. The analysis involved decontextualization, recontextualization, categorization, and compilation. We primarily developed inductive categories, which were subsequently applied deductively. After completion, we performed secondary deductive coding using the following six stages of the development and course of second victim effects proposed by Scott as an established framework for second victim evaluation: “(1) chaos and accident response (2), intrusive reflections (3), restoring personal integrity (4), enduring the inquisition (5), obtaining emotional first aid, and (6) moving on (drop-out, survive, thrive)” [12]. Our steps were accomplished iteratively by a single coder (SB), with an independent peer review by two secondary coders (SBu and TT). The main male coder and initiator of the study worked as an emergency and intensive care physician, infection control specialist, and FEES trainer in two primary hospitals and one children’s rehabilitation hospital with 17 years of experience in intensive care. The second coder was a female trauma surgeon and emergency physician in a rehabilitation hospital with 16 years of experience in emergency medicine and traumatology. Both coders have regularly worked with patients with a tracheostomy and dysphagia for three (SBu)and 13 years (SB). The third coder was a male undergraduate medical student (5th year) and an emergency medical technician (EMT; basic) without experience with dysphagia treatment, providing an external perspective on the free-text entries. Codes without agreement between the coders were discussed iteratively and adapted if at least two coders agreed. Contact or feedback was not possible between the participants and coders. As some participants responded with only single words and “bullet points,” the smallest meaning unit analyzed was words. We used a codebook and logbook to control for coder drift and cognitive changes. We began an iterative coding process for different cases to maintain process stability. We used MAXQDA (MAXQDA, Software für qualitative Datenanalyse, 1989–2024, VERBI Software. Consult. Sozialforschung GmbH, Berlin, Germany) for code compilation, code tree generation, and summaries for coding (e.g., code matrices). We identified both manifest qualitative (phrases, citations) and quantitative codes (counts, percentages) and added impressions of positive or negative valence. Codes were not interpreted as a proportional indicator for “proof” or the “weight” of an argument. Every code, even if mentioned only once, is considered potentially meaningful, particularly for hypothesis generation. Additionally, we used the MAXQDA Matrix Browser to support structural analysis of the content and orientation for pattern recognition and relationships between and co-occurrence of codes in the same paragraph (“proximity codes”). The code count was relevant for secondary analysis. We used DeepL (DeepL SE, Cologne, Germany) to translate German free-text entries with a native speaker, thus confirming the idiomatic meaning. Details of the codes and themes are provided in the Supplementary Materials. Results Participants A total of 158 participants agreed to participate and were involved in the study. Prior to this, we excluded three persons because they did not agree to their data being used. Of the 158 participants, 94 (59.5%) completed the survey. With a focus on the primary results, we excluded 31 participants who did not respond beyond the demographics section, preserving 127 meaningful responses (33 “dropouts” and 94 “finishers”). The demographics of all respondents are summarized in Table 1 and Fig. 1. Table 1Categorical demographics of the 127 respondents. 1) many free-text entries were obtained from professionals. Speech and language therapists have several specializations and academic profiles (e.g., patho-linguists, clinical linguists, neurolinguists, speech therapists); we categorized these persons into speech and language therapists and respiratory therapists. 2) one responder working for years in Germany, but who moved to Switzerland a few months ago, was accounted for as in GermanyCategoricalparametersN (of 127)%SexFemaleMaleNon-binaryNo answer111131287.410.20.81.6Professional group1Speech language therapistsNurse respiratory therapistsOthers/not specified11014386.611.02.4OccupationalsettingHospitalRehabilitation hospitalOut of hospital, mixed, not specified43622233.948.817.3Work countryGermany2AustriaSwitzerland1157590.65.53.9ContinuousparametersMinMaxMeanSDMedianIQRAge216140.811.34019.5Estimatedpercentage of patients withdysphagiacurrently treated010069.425.47040Estimatedpercentage of patientswith atracheostomycurrently treated010039.134.13060Fig. 1Study flowchart Table 1Categorical demographics of the 127 respondents. 1) many free-text entries were obtained from professionals. Speech and language therapists have several specializations and academic profiles (e.g., patho-linguists, clinical linguists, neurolinguists, speech therapists); we categorized these persons into speech and language therapists and respiratory therapists. 2) one responder working for years in Germany, but who moved to Switzerland a few months ago, was accounted for as in GermanyCategoricalparametersN (of 127)%SexFemaleMaleNon-binaryNo answer111131287.410.20.81.6Professional group1Speech language therapistsNurse respiratory therapistsOthers/not specified11014386.611.02.4OccupationalsettingHospitalRehabilitation hospitalOut of hospital, mixed, not specified43622233.948.817.3Work countryGermany2AustriaSwitzerland1157590.65.53.9ContinuousparametersMinMaxMeanSDMedianIQRAge216140.811.34019.5Estimatedpercentage of patients withdysphagiacurrently treated010069.425.47040Estimatedpercentage of patientswith atracheostomycurrently treated010039.134.13060 Categorical demographics of the 127 respondents. 1) many free-text entries were obtained from professionals. Speech and language therapists have several specializations and academic profiles (e.g., patho-linguists, clinical linguists, neurolinguists, speech therapists); we categorized these persons into speech and language therapists and respiratory therapists. 2) one responder working for years in Germany, but who moved to Switzerland a few months ago, was accounted for as in Germany Categorical parameters Female Male Non-binary No answer 111 13 1 2 87.4 10.2 0.8 1.6 Speech language therapists Nurse respiratory therapists Others/not specified 110 14 3 86.6 11.0 2.4 Occupational setting Hospital Rehabilitation hospital Out of hospital, mixed, not specified 43 62 22 33.9 48.8 17.3 Germany2 Austria Switzerland 115 7 5 90.6 5.5 3.9 Continuous parameters Estimated percentage of patients with dysphagia currently treated Estimated percentage of patients with a tracheostomy currently treated Study flowchart By inspecting the histograms, we found an approximately normal distribution for age but not for the percentage of patients with dysphagia (left-skewed, many with high percentages) and a tracheal cannula (right-skewed, zero-inflation, many without a tracheostomy). Treatments of patients with dysphagia and a tracheostomy were positively correlated (see Fig. 2). Fig. 2Demographical overview: scatterplot of participants’ self-reported exposure to patients with dysphagia and a tracheostomy. Treating patients with dysphagia is highly correlated with treating those with a tracheostomy (Spearman’s rho = 0.7). SLT = speech and language therapist and aligned academic and non-academic professions, RLT = registered respiratory therapists, OTHER = others (e.g., physicians, occupational therapists, or nurses not specialized in respiratory therapy) Demographical overview: scatterplot of participants’ self-reported exposure to patients with dysphagia and a tracheostomy. Treating patients with dysphagia is highly correlated with treating those with a tracheostomy (Spearman’s rho = 0.7). SLT = speech and language therapist and aligned academic and non-academic professions, RLT = registered respiratory therapists, OTHER = others (e.g., physicians, occupational therapists, or nurses not specialized in respiratory therapy) Self-reported experience of SVP Overall, 71.65% (91 of 127) of the participants reported having experienced a second victim effect at least once in their lifetime and 39.4% (50 of 127) reported experiencing it more than once. Twenty of the 127 participants (15.8%) knew the term second victim before this investigation (see Table 2). Table 2Second victim experiences. Information derived from questions on whether they have been affected by a second victim effect or not. Percentages are computed per groupAllNoSVEreported(n = 36)OneSVE reported(n = 41)MultipleSVEs reported(n = 50)All SVEsreported (n = 91)Nn%n%n%N%Second Victim Effect ExperiencesKnowledge of the second victim effect12738.3614.61122.01718.7Prolonged experience (>1 year)12112.8512.2714.01220.1No complete recovery reported11925.6819.51224.02022.0Association with dysphagia/tracheostomy1271130.62970.74182.07076.9Help was sought for the situation114616.71434.11632.03033.0 Table 2Second victim experiences. Information derived from questions on whether they have been affected by a second victim effect or not. Percentages are computed per groupAllNoSVEreported(n = 36)OneSVE reported(n = 41)MultipleSVEs reported(n = 50)All SVEsreported (n = 91)Nn%n%n%N%Second Victim Effect ExperiencesKnowledge of the second victim effect12738.3614.61122.01718.7Prolonged experience (>1 year)12112.8512.2714.01220.1No complete recovery reported11925.6819.51224.02022.0Association with dysphagia/tracheostomy1271130.62970.74182.07076.9Help was sought for the situation114616.71434.11632.03033.0 Second victim experiences. Information derived from questions on whether they have been affected by a second victim effect or not. Percentages are computed per group Among the 91 affected persons, 12 reported being affected for longer than one year (13.2%). Twenty of them (22%) did not feel that they had recovered from the incident. Further, 77% of the reported second victim effects were aligned with dysphagia or tracheostomy treatment (70/91). Of those reporting that they were not affected, six of 36 (16.7%) participants sought help after a related incident. Further, 30 of the 91 second victims (33.0%) reported seeking help after an incident. Concerning “symptoms,” 91 respondents identifying themselves as second victims reported the high prevalence of self-doubt (72.5%), guilt (61.5%), a desire to analyze the event (50.5%), defensive behavior (41.8%), and re-experiencing the situation (35.2%). For an explorative assessment, we also report the findings for persons not feeling affected but showing similar symptoms, including guilt, self-doubt, and the desire for support. Only half of them (18/36, 50%) denied having symptoms and confirmed a second time that there was no such event. Further results for subjective experiences and associations with dysphagia and respiratory care events are shown in Tables 3 and 4. Table 3General context and situation of the eventNoSVEReported(n = 36)OneSVE reported(n = 41)MultipleSVEsreported(n = 50)AllSVEs reported(n = 91)General Context and Situation of the Incidentn/36%n/41%n/50%n/91%Adverse event with patient harm513.9%1126.8%1938.0%3033.0%Adverse event without patient harm616.7%1843.9%3162.0%4953.8%Unexpected death/suicide(patient)25.6%1434.1%1632.0%3033.0%Unexpected death/suicide (colleague)00.0%00.0%12.0%11.1%Aggressive behavior(patient, family member)513.9%614.6%510.0%1112.1%Table 4Specific context and situation of the event. Further single mentions were as follows: error by a physician, death of a patient during FEES without information to the therapist, own error in tracheal cannula management, change of tracheal cannula, syncope triggered by oral suctioning, patient wanted to eat despite severe dysphagia, reinsertion of tracheal cannula failed due to a fistula, and occlusion cap use for tracheal cannulaSpecific Context and Situation of the IncidentNoSVEReported(n = 36)OneSVE reported(n = 41)MultipleSVEsreported(n = 50)AllSVEs reported(n = 91)Critical eventn/36%n/41%n/50%n/91%Dislocation of a tracheal cannula00.0%37.3%510.0%88.8%Bleeding from tracheostomy12.8%37.3%510.0%88.8%Decannulation25.6%49.8%510.0%99.9%Endotracheal suctioning12.8%37.3%36.0%66.6%Cuff deflation (e.g., with hypoxygenation)38.3%922.0%1734.0%2628.6%Problems with the ventilator machine00.0%24.9%24.0%44.4%Severe aspiration after a swallowing test12.8%12.4%816.0%99.9%Severe aspiration after feeding advice25.6%717.1%816.0%1516.5%Loss of weight due to food restriction00.0%12.4%00.0%11.1%Loss of quality of life due to food restriction25.6%24.9%816.0%1011.0%Risk after experiencing pressure fromfamily members that the patient should eat “normally”25.6%24.9%1122.0%1314.3%Pneumonia38.3%1024.4%1224.0%2224.2%Sepsis00.0%12.4%24.0%33.3%Choking00.0%24.9%48.0%66.6%Cardiopulmonary resuscitation12.8%512.2%510.0%1011.0%Palliative situation12.8%37.3%714.0%1011.0%Psychological crisis (patient)00.0%12.4%48.0%55.5%Psychological crisis (family member)00.0%12.4%24.0%33.3%Fiberoptic evaluation of swallowing (FEES)00.0%24.9%36.0%55.5%Conflict with caregivers00.0%37.3%36.0%66.6%Conflict with patients00.0%24.9%48.0%66.6%Conflict with doctors12.8%512.2%36.0%88.8%Legal issues00.0%00.0%00.0%00.0%Denial of the event12.8%12.4%00.0%11.1% Table 3General context and situation of the eventNoSVEReported(n = 36)OneSVE reported(n = 41)MultipleSVEsreported(n = 50)AllSVEs reported(n = 91)General Context and Situation of the Incidentn/36%n/41%n/50%n/91%Adverse event with patient harm513.9%1126.8%1938.0%3033.0%Adverse event without patient harm616.7%1843.9%3162.0%4953.8%Unexpected death/suicide(patient)25.6%1434.1%1632.0%3033.0%Unexpected death/suicide (colleague)00.0%00.0%12.0%11.1%Aggressive behavior(patient, family member)513.9%614.6%510.0%1112.1% General context and situation of the event Unexpected death/suicide (patient) Aggressive behavior (patient, family member) Table 4Specific context and situation of the event. Further single mentions were as follows: error by a physician, death of a patient during FEES without information to the therapist, own error in tracheal cannula management, change of tracheal cannula, syncope triggered by oral suctioning, patient wanted to eat despite severe dysphagia, reinsertion of tracheal cannula failed due to a fistula, and occlusion cap use for tracheal cannulaSpecific Context and Situation of the IncidentNoSVEReported(n = 36)OneSVE reported(n = 41)MultipleSVEsreported(n = 50)AllSVEs reported(n = 91)Critical eventn/36%n/41%n/50%n/91%Dislocation of a tracheal cannula00.0%37.3%510.0%88.8%Bleeding from tracheostomy12.8%37.3%510.0%88.8%Decannulation25.6%49.8%510.0%99.9%Endotracheal suctioning12.8%37.3%36.0%66.6%Cuff deflation (e.g., with hypoxygenation)38.3%922.0%1734.0%2628.6%Problems with the ventilator machine00.0%24.9%24.0%44.4%Severe aspiration after a swallowing test12.8%12.4%816.0%99.9%Severe aspiration after feeding advice25.6%717.1%816.0%1516.5%Loss of weight due to food restriction00.0%12.4%00.0%11.1%Loss of quality of life due to food restriction25.6%24.9%816.0%1011.0%Risk after experiencing pressure fromfamily members that the patient should eat “normally”25.6%24.9%1122.0%1314.3%Pneumonia38.3%1024.4%1224.0%2224.2%Sepsis00.0%12.4%24.0%33.3%Choking00.0%24.9%48.0%66.6%Cardiopulmonary resuscitation12.8%512.2%510.0%1011.0%Palliative situation12.8%37.3%714.0%1011.0%Psychological crisis (patient)00.0%12.4%48.0%55.5%Psychological crisis (family member)00.0%12.4%24.0%33.3%Fiberoptic evaluation of swallowing (FEES)00.0%24.9%36.0%55.5%Conflict with caregivers00.0%37.3%36.0%66.6%Conflict with patients00.0%24.9%48.0%66.6%Conflict with doctors12.8%512.2%36.0%88.8%Legal issues00.0%00.0%00.0%00.0%Denial of the event12.8%12.4%00.0%11.1% Specific context and situation of the event. Further single mentions were as follows: error by a physician, death of a patient during FEES without information to the therapist, own error in tracheal cannula management, change of tracheal cannula, syncope triggered by oral suctioning, patient wanted to eat despite severe dysphagia, reinsertion of tracheal cannula failed due to a fistula, and occlusion cap use for tracheal cannula Risk after experiencing pressure from family members that the patient should eat “normally” Although 36 participants initially reported that they had not experienced an SVP in the first section (asking about symptoms), only 18 (50%) actively denied symptoms, while others reported event-related symptoms (Table 5). In the second section (asking about severe adverse events), only one person of those not identifying as second victims initially, actively stated for a third time that there was no event. The remaining participants, not initially identifying as second victims, reported up to eight incident-related symptoms, including depression, flashbacks, sleep disorders, and alcohol use for relief. Another person did not report an event but reported identifying as a second victim in one case with complete recovery. Table 5Symptoms after an adverse or critical event. Self-reported symptoms after a critical event that align with the definition of a second victim effect. Again, this question was not gated by self-reported responses of being affected or not. Percentages are calculated based on the number of mentions in the same group. Data for the “all SVEs” group presents a summary of those reporting having been affected by a second victim effect once or multiple timesNoSVEreported(n = 36)OneSVE reported(n = 41)MultipleSVEsreported(n = 50)AllSVEs reported(n = 91)Symptoms After Adverse Events and Second Victim Experiencesn%n%n%n%Fear of being avoided by colleagues12.8%1126.8%1122.0%2224.2%Fear of losing the job12.8%49.8%714.0%1112.1%Listlessness00.0%37.3%612.0%99.9%Depressive mood25.6%49.8%1734.0%2123.1%Concentration difficulties25.6%49.8%1326.0%1718.7%“Flashbacks”/re-experience (out-of-the-job situation)25.6%1024.4%1122.0%2123.1%“Flashbacks”/re-experience (in-job situation)12.8%1126.8%2142.0%3235.2%Aggressive and risky behavior25.6%00.0%00.0%00.0%Defensive behavior38.3%1843.9%2040.0%3841.8%Psychosomatic reactions (e.g., headache, back pain)12.8%1126.8%1428.0%2527.5%Sleep disorders411.1%1126.8%1938.0%3033.0%Alcohol and drug consumption12.8%12.4%48.0%55.5%Feelings of shame411.1%614.6%1734.0%2325.3%Feelings of guilt616.7%2151.2%3570.0%5661.5%Self-doubt616.7%2970.7%3774.0%6672.5%Social isolation12.8%00.0%36.0%33.3%Anger toward others513.9%819.5%1428.0%2224.2%Anger toward the self12.8%922.0%1428.0%2325.3%Desire for support by others822.2%1331.7%2244.0%3538.5%Desire to analyze and understand the event411.1%1536.6%3162.0%4650.5%No symptoms, there was no event1850.0%00.0%00.0%00.0% Table 5Symptoms after an adverse or critical event. Self-reported symptoms after a critical event that align with the definition of a second victim effect. Again, this question was not gated by self-reported responses of being affected or not. Percentages are calculated based on the number of mentions in the same group. Data for the “all SVEs” group presents a summary of those reporting having been affected by a second victim effect once or multiple timesNoSVEreported(n = 36)OneSVE reported(n = 41)MultipleSVEsreported(n = 50)AllSVEs reported(n = 91)Symptoms After Adverse Events and Second Victim Experiencesn%n%n%n%Fear of being avoided by colleagues12.8%1126.8%1122.0%2224.2%Fear of losing the job12.8%49.8%714.0%1112.1%Listlessness00.0%37.3%612.0%99.9%Depressive mood25.6%49.8%1734.0%2123.1%Concentration difficulties25.6%49.8%1326.0%1718.7%“Flashbacks”/re-experience (out-of-the-job situation)25.6%1024.4%1122.0%2123.1%“Flashbacks”/re-experience (in-job situation)12.8%1126.8%2142.0%3235.2%Aggressive and risky behavior25.6%00.0%00.0%00.0%Defensive behavior38.3%1843.9%2040.0%3841.8%Psychosomatic reactions (e.g., headache, back pain)12.8%1126.8%1428.0%2527.5%Sleep disorders411.1%1126.8%1938.0%3033.0%Alcohol and drug consumption12.8%12.4%48.0%55.5%Feelings of shame411.1%614.6%1734.0%2325.3%Feelings of guilt616.7%2151.2%3570.0%5661.5%Self-doubt616.7%2970.7%3774.0%6672.5%Social isolation12.8%00.0%36.0%33.3%Anger toward others513.9%819.5%1428.0%2224.2%Anger toward the self12.8%922.0%1428.0%2325.3%Desire for support by others822.2%1331.7%2244.0%3538.5%Desire to analyze and understand the event411.1%1536.6%3162.0%4650.5%No symptoms, there was no event1850.0%00.0%00.0%00.0% Symptoms after an adverse or critical event. Self-reported symptoms after a critical event that align with the definition of a second victim effect. Again, this question was not gated by self-reported responses of being affected or not. Percentages are calculated based on the number of mentions in the same group. Data for the “all SVEs” group presents a summary of those reporting having been affected by a second victim effect once or multiple times Explorative regression analysis to determine predictors of second victim effect self-reporting The binomial logistic regression model operationalized the second victim experience as an outcome, and age, sex, and BFI-10 personality traits as predictors. We used data from 124 persons, with three datasets deleted as mentioned above. The model was not statistically signiant, χ2(7) = 8.61, p = 0.28, residual-df = 116. This result indicates that the model did not provide a better fit than the null model. The explained variance was low, with a Nagelkerke’s pseudo-R2 value of 0.096. The accuracy was 0.73, sensitivity was 0.97, specificity was 0.11, and precision was 0.74, indicating an underestimation of negative cases. Among all predictors, neuroticism was the only variable showing a statistically significant effect (β = 0.55, SE = 0.27, p = 0.045; OR = 1.74), suggesting a higher likelihood of the outcome for individuals with higher scores in neuroticism. The R-bootstrapped 95% confidence interval for the odds ratio did not confirm the effect (BCa 95%, 5000 samples, CIOR: [0.94; 3.16]). Multicollinearity was not an issue (VIF < 5 for all predictors). Further details are presented in Tables S1−S5 in the Supplementary Materials. Content analysis of free-text entries to generate new hypotheses The analyzed 76 free-text entries for questions 19 and 20 and the final question were related to 10 persons who did not reported being a second victim and 66 who did. We used the following classifications for those formulating free-text entries: “self-aware” second victims (n = 66), with 31 experiencing one event and 35 experiencing multiple events; and “non-victims” (n = 10). Of the 10 “non-victims,” nine persons (90%) reported situations, support, or outcomes related to adverse events, contradicting their initial self-reports. As these events all fall under the definition of the second victim effect [10], we categorized these individuals as “silent second victims.” During the text coding process, we noted and discussed the fact that these codes formed themes along the following described pathway of a second victim effect: predispositions, the event with modulators, a first reaction, first self-help, support from others, support from psychology professionals, an outcome, and self-identification as a second victim or not [16]. Table 6 provides an overview of the themes and codes along with the corresponding references to the literature. In addition, we implemented the measurement of “proximity codes” using the MAXQDA code-relationship-browser. This included the detection of code relationships and co-occurrences that were repeatedly detected in the text within the same paragraph as another code. The Supplementary Materials provide detailed descriptions, code derivations, and further references. Table 6Codes and themes, quantification of inductive codes, and qualitative analysis of free-text entries. Proximity codes repeatedly occur near a code according to the MAXQDA matrix function, meaning that this information is often mentioned togetherInductive codesSubcode(n)MeaningEvidence1. Theme 1: predisposition to second victim experiences1.1 Institutional readinessPreparedness of an institution or organization to manage adverse events and SVPEpidemiology of events5Frequency of adverse events happeningProximity codes: Safety culture (10)[38]Building infrastructure1Separated rooms for rest and communicationProximity codes: -[39]Interprofessional care7SVP interaction with different professions and disciplinesProximity codes: Roles and responsibilities (13)[40]Roles and responsibilities21SVP occurrence and interactions with the responsibilities of different healthcare providersProximity codes: Interprofessional care (13), safety culture (18), qualitative staff resources (12), disappointment (14), guilt (14), and rationalization (13)[41, 42]+Safety culture12Predisposition depends on the safety culture level of an institutionProximity codes: Roles and responsibilities (18), rationalization (11), hot and cold debriefing (13), communication peers (14)[43, 44]Staff resources (quantitative)6SVP phenomena due to staff shortagesProximity codes: Roles and responsibilities (12)[45]Staff resources (qualitative/educative)5SVP phenomena due to educational gaps and misqualificationProximity codes: -1.2 Suprainstitutional factorsFactors attributable to the whole health system instead to institutions or organizationsHealthcare system3National factors, laws, regulations, and othersProximity codes: Roles and responsibilities (7)Not applicableGender awareness30Detection of (primarily linguistic) gender diversity aspectsProximity codes: Roles and responsibilities (21)Not applicable1.3 Individual factorsIndividual predisposing factorsMoral distress and injury3Factors of predisposition by moral distress and moral injuryProximity codes: Roles and responsibilities (21)[34]Theme 2: Events leading to second victim phenomena2.1 Medical causesMedical causes of SVPDeath and peri-arrest4Death or life-threatening situation (e.g., shock, myocardial infarction, sepsis)Proximity codes: -[17]Nutrition1Nutritional situation (e.g., starving, cachexia)Proximity codes: -*Airway6Issues with the airway and airway devices (choking, tracheal cannula problems)Proximity codes: -[46]2.2 Event modulatorModulatory events that provoke or attenuate traumatic eventsConflicts1Conflicts with other persons in a stressful situationProximity codes: -[45]Emotional binding1Emotional binding to a patient experiencing a difficult situationProximity codes: -[21]Helplessness3Situations of feeling overwhelmed or helplessProximity codes: -[24]Near miss1A near mistake without any harmProximity codes: -[19]Routine situation2A difficult, unexpected situation in routine careProximity codes: -[24]Theme 3: Reactions to second victim experiences3.1 PhysiologicalPhysiological reactionsPain1Feeling of painProximity codes: -[17, 19]Intrusive thoughts1Recurrent phases of remembering the situationProximity codes: -[12]3.2. PsychologicalPsychological reactionsDisappointment5Feeling disappointed by the situationProximity codes: Roles and responsibilities (13), safety culture (7)[34, 47]Interest15Feeling interested in the situation or phenomenon, information seeking.Proximity codes: Roles and responsibilities (9), rationalization (13), hot and cold debriefing (14), communication with peers (23), thriving (13), “multiple” victim (18)[30]Shame2Feeling ashamed of one’s own decisions or actionsProximity codes: -[14]Guilt11Feeling guilty for the adverse eventProximity codes: Roles and responsibilities (14), safety culture (10), restoring personal integrity (11), rationalization (9), thriving (10)[14]3.3 Emotion“Semi”-deductive code for emotional reactions that is part of the Ekman model[48]Fear7Phrases expressing fearProximity codes: Roles and responsibilities (13), safety culture (10), rationalization (9), knowledge and learning (7), thriving (10), intrusive reflections (9), restoring personal integrity (11)[14]Anger3Phrases expressing angerProximity codes: -[14]Enjoyment2Phrases expressing enjoymentProximity codes: -*Disgust1Phrases expressing disgustProximity codes: -[24]Sadness3Phrases expressing sadnessProximity codes: -[24]Theme 4: Support after a traumatic event4.1. Self-supportAll kinds of self-support after an event4.1.1 Reflexive Self-supportInternal reflection without external informationRepression4Repression of the situationProximity codes: -*Self-justification2Justification of one’s own role in the eventProximity codes: -[12]Rationalization15Trying to understand for oneself what and how it happenedProximity codes: Roles and responsibilities (13), safety culture (11), restoring personal integrity (17), interest (13), knowledge and learning (9), fear (9)[30, 49, 50]4.1.2 Generative self-supportActive actions initiated by oneself, but not an intervention by othersKnowledge and learning14Retrieving knowledge about the events and building competenceProximity codes: Restoring personal integrity (11), rationalization (9), passing time (8), peer communication (13), family support (11), thriving (21)[12]Teaching1Teaching others about the event to prevent them from making the same mistakeProximity codes: -*Mindfulness2Being mindful of oneselfProximity codes: -[30, 49, 51]4.1.3 AbsenteeismCodes on being present or absent at the workplaceAbsenteeism (rest)1Being absent from the situation and taking restProximity codes: -[14, 30, 49]Absenteeism (others)3Being absent to avoid situations or personsProximity codes: -[14, 30, 49]4.1.4 PresenteeismCodes on being present at the workplacePresenteeism (structures and routines)2Being present to follow workplace structures and daily routinesProximity codes: -[30, 31]4.1.5 Physical self-supportCoping mechanism after an eventSports and activity3Physical activitiesProximity codes: -[52]Passing time14“Time heals all wounds”Proximity codes: -[53]Alcohol1Alcohol for relief after an event (NOT alcoholism)Proximity codes: -4.2. Leadership supportCommunication to supervisors and the management, leading to structured discussion and changeInstitutional changes in patient safety5Coded for changes in an organization or institution after an event, preventing the event from occurring again, or reducing the riskProximity codes: Restoring personal integrity (8), roles and responsibilities (6), patient safety (7), airway (6), thriving (10)[54]Hot and cold debriefing16Performing a direct (hot) debriefing after an event or a structural debriefing days later (cold)Proximity codes: Roles and responsibilities (10), patient safety (13), restoring personal integrity (16), emotional first aid (14), interest (14), rationalization (13), communication with peers (25), thriving (10)[55]Individual communication with a supervisor4Communicating with the supervisor about the eventProximity codes: -[30, 49]4.3. Peer supportSupport from persons in the same hierarchy level (e.g., colleagues)Interprofessional communication7Peer support from persons of another discipline or professionProximity codes: Roles and responsibilities (7), interprofessional care (5), emotional first aid (6), professional therapy (7)[56]Communication (any)10Peer support without any further definitionProximity codes: Emotional first aid (16)[56]Communication (peers)32Peer support from persons of the same discipline or professionProximity codes: Roles and responsibilities (14), patient safety (14), restoring personal integrity (21), emotional first aid (33), thriving (14), interest (23), fear (8), rationalization (17), knowledge and learning (13), passing time (13), hot and cold debriefing (25), family support (12)[56]4.4. Family support9Support from family membersProximity codes: Knowledge and learning (9), restoring personal integrity (8), emotional first aid (10), communication (peers) (12), thriving (8)[30]4.5. Support from first victims2Support from patients and proxies who were involved or harmedProximity codes: -*4.6. Support from professionals8All support is provided by professionals who are educated to deal with second victims and traumatized personsProximity codes: Roles and responsibilities (6), emotional first aid (6), “multiple” victim (10), drop-outs (2), survivors (4), unspecified outcome (9)[56]Theme 5: Outcome and classification5.1. OutcomeNo victimization1Not a second victimProximity codes: -[12]Surviving10Dysfunctional behavior, but staying in placeProximity codes: -[12]Drop-out3Dropping out of the workplace after eventsProximity codes: -[12]Thriving21Growing from the eventProximity codes: -[12]Unknown outcome54Unknown outcomeProximity codes: -Not applicable5.2. ClassificationNon-victim1A person actively denying being a second victim multiple timesProximity codes: -Mono-victim31Persons reporting one eventProximity codes: -Multi-victim35Persons reporting more than one eventProximity codes: -Silent second victim9Persons not identifying as a second victim but reporting events and traumatization according to the second victim definitionProximity codes: -*Recall bias4Persons reporting that they had forgotten about their second victim experience and realized it during the surveyProximity codes: -*The evidence column shows exemplary articles and manuscripts that described the “code” earlier. Codes with an asterisk (+) are “central” codes that show many relationships to other codes and subcodes in the code-browser-matrix. These proximity codes are labeled with a number, showing the number of the linked code in proximity to the described code. Codes with an * are those that could potentially inform new hypotheses Table 6Codes and themes, quantification of inductive codes, and qualitative analysis of free-text entries. Proximity codes repeatedly occur near a code according to the MAXQDA matrix function, meaning that this information is often mentioned togetherInductive codesSubcode(n)MeaningEvidence1. Theme 1: predisposition to second victim experiences1.1 Institutional readinessPreparedness of an institution or organization to manage adverse events and SVPEpidemiology of events5Frequency of adverse events happeningProximity codes: Safety culture (10)[38]Building infrastructure1Separated rooms for rest and communicationProximity codes: -[39]Interprofessional care7SVP interaction with different professions and disciplinesProximity codes: Roles and responsibilities (13)[40]Roles and responsibilities21SVP occurrence and interactions with the responsibilities of different healthcare providersProximity codes: Interprofessional care (13), safety culture (18), qualitative staff resources (12), disappointment (14), guilt (14), and rationalization (13)[41, 42]+Safety culture12Predisposition depends on the safety culture level of an institutionProximity codes: Roles and responsibilities (18), rationalization (11), hot and cold debriefing (13), communication peers (14)[43, 44]Staff resources (quantitative)6SVP phenomena due to staff shortagesProximity codes: Roles and responsibilities (12)[45]Staff resources (qualitative/educative)5SVP phenomena due to educational gaps and misqualificationProximity codes: -1.2 Suprainstitutional factorsFactors attributable to the whole health system instead to institutions or organizationsHealthcare system3National factors, laws, regulations, and othersProximity codes: Roles and responsibilities (7)Not applicableGender awareness30Detection of (primarily linguistic) gender diversity aspectsProximity codes: Roles and responsibilities (21)Not applicable1.3 Individual factorsIndividual predisposing factorsMoral distress and injury3Factors of predisposition by moral distress and moral injuryProximity codes: Roles and responsibilities (21)[34]Theme 2: Events leading to second victim phenomena2.1 Medical causesMedical causes of SVPDeath and peri-arrest4Death or life-threatening situation (e.g., shock, myocardial infarction, sepsis)Proximity codes: -[17]Nutrition1Nutritional situation (e.g., starving, cachexia)Proximity codes: -*Airway6Issues with the airway and airway devices (choking, tracheal cannula problems)Proximity codes: -[46]2.2 Event modulatorModulatory events that provoke or attenuate traumatic eventsConflicts1Conflicts with other persons in a stressful situationProximity codes: -[45]Emotional binding1Emotional binding to a patient experiencing a difficult situationProximity codes: -[21]Helplessness3Situations of feeling overwhelmed or helplessProximity codes: -[24]Near miss1A near mistake without any harmProximity codes: -[19]Routine situation2A difficult, unexpected situation in routine careProximity codes: -[24]Theme 3: Reactions to second victim experiences3.1 PhysiologicalPhysiological reactionsPain1Feeling of painProximity codes: -[17, 19]Intrusive thoughts1Recurrent phases of remembering the situationProximity codes: -[12]3.2. PsychologicalPsychological reactionsDisappointment5Feeling disappointed by the situationProximity codes: Roles and responsibilities (13), safety culture (7)[34, 47]Interest15Feeling interested in the situation or phenomenon, information seeking.Proximity codes: Roles and responsibilities (9), rationalization (13), hot and cold debriefing (14), communication with peers (23), thriving (13), “multiple” victim (18)[30]Shame2Feeling ashamed of one’s own decisions or actionsProximity codes: -[14]Guilt11Feeling guilty for the adverse eventProximity codes: Roles and responsibilities (14), safety culture (10), restoring personal integrity (11), rationalization (9), thriving (10)[14]3.3 Emotion“Semi”-deductive code for emotional reactions that is part of the Ekman model[48]Fear7Phrases expressing fearProximity codes: Roles and responsibilities (13), safety culture (10), rationalization (9), knowledge and learning (7), thriving (10), intrusive reflections (9), restoring personal integrity (11)[14]Anger3Phrases expressing angerProximity codes: -[14]Enjoyment2Phrases expressing enjoymentProximity codes: -*Disgust1Phrases expressing disgustProximity codes: -[24]Sadness3Phrases expressing sadnessProximity codes: -[24]Theme 4: Support after a traumatic event4.1. Self-supportAll kinds of self-support after an event4.1.1 Reflexive Self-supportInternal reflection without external informationRepression4Repression of the situationProximity codes: -*Self-justification2Justification of one’s own role in the eventProximity codes: -[12]Rationalization15Trying to understand for oneself what and how it happenedProximity codes: Roles and responsibilities (13), safety culture (11), restoring personal integrity (17), interest (13), knowledge and learning (9), fear (9)[30, 49, 50]4.1.2 Generative self-supportActive actions initiated by oneself, but not an intervention by othersKnowledge and learning14Retrieving knowledge about the events and building competenceProximity codes: Restoring personal integrity (11), rationalization (9), passing time (8), peer communication (13), family support (11), thriving (21)[12]Teaching1Teaching others about the event to prevent them from making the same mistakeProximity codes: -*Mindfulness2Being mindful of oneselfProximity codes: -[30, 49, 51]4.1.3 AbsenteeismCodes on being present or absent at the workplaceAbsenteeism (rest)1Being absent from the situation and taking restProximity codes: -[14, 30, 49]Absenteeism (others)3Being absent to avoid situations or personsProximity codes: -[14, 30, 49]4.1.4 PresenteeismCodes on being present at the workplacePresenteeism (structures and routines)2Being present to follow workplace structures and daily routinesProximity codes: -[30, 31]4.1.5 Physical self-supportCoping mechanism after an eventSports and activity3Physical activitiesProximity codes: -[52]Passing time14“Time heals all wounds”Proximity codes: -[53]Alcohol1Alcohol for relief after an event (NOT alcoholism)Proximity codes: -4.2. Leadership supportCommunication to supervisors and the management, leading to structured discussion and changeInstitutional changes in patient safety5Coded for changes in an organization or institution after an event, preventing the event from occurring again, or reducing the riskProximity codes: Restoring personal integrity (8), roles and responsibilities (6), patient safety (7), airway (6), thriving (10)[54]Hot and cold debriefing16Performing a direct (hot) debriefing after an event or a structural debriefing days later (cold)Proximity codes: Roles and responsibilities (10), patient safety (13), restoring personal integrity (16), emotional first aid (14), interest (14), rationalization (13), communication with peers (25), thriving (10)[55]Individual communication with a supervisor4Communicating with the supervisor about the eventProximity codes: -[30, 49]4.3. Peer supportSupport from persons in the same hierarchy level (e.g., colleagues)Interprofessional communication7Peer support from persons of another discipline or professionProximity codes: Roles and responsibilities (7), interprofessional care (5), emotional first aid (6), professional therapy (7)[56]Communication (any)10Peer support without any further definitionProximity codes: Emotional first aid (16)[56]Communication (peers)32Peer support from persons of the same discipline or professionProximity codes: Roles and responsibilities (14), patient safety (14), restoring personal integrity (21), emotional first aid (33), thriving (14), interest (23), fear (8), rationalization (17), knowledge and learning (13), passing time (13), hot and cold debriefing (25), family support (12)[56]4.4. Family support9Support from family membersProximity codes: Knowledge and learning (9), restoring personal integrity (8), emotional first aid (10), communication (peers) (12), thriving (8)[30]4.5. Support from first victims2Support from patients and proxies who were involved or harmedProximity codes: -*4.6. Support from professionals8All support is provided by professionals who are educated to deal with second victims and traumatized personsProximity codes: Roles and responsibilities (6), emotional first aid (6), “multiple” victim (10), drop-outs (2), survivors (4), unspecified outcome (9)[56]Theme 5: Outcome and classification5.1. OutcomeNo victimization1Not a second victimProximity codes: -[12]Surviving10Dysfunctional behavior, but staying in placeProximity codes: -[12]Drop-out3Dropping out of the workplace after eventsProximity codes: -[12]Thriving21Growing from the eventProximity codes: -[12]Unknown outcome54Unknown outcomeProximity codes: -Not applicable5.2. ClassificationNon-victim1A person actively denying being a second victim multiple timesProximity codes: -Mono-victim31Persons reporting one eventProximity codes: -Multi-victim35Persons reporting more than one eventProximity codes: -Silent second victim9Persons not identifying as a second victim but reporting events and traumatization according to the second victim definitionProximity codes: -*Recall bias4Persons reporting that they had forgotten about their second victim experience and realized it during the surveyProximity codes: -*The evidence column shows exemplary articles and manuscripts that described the “code” earlier. Codes with an asterisk (+) are “central” codes that show many relationships to other codes and subcodes in the code-browser-matrix. These proximity codes are labeled with a number, showing the number of the linked code in proximity to the described code. Codes with an * are those that could potentially inform new hypotheses Codes and themes, quantification of inductive codes, and qualitative analysis of free-text entries. Proximity codes repeatedly occur near a code according to the MAXQDA matrix function, meaning that this information is often mentioned together Frequency of adverse events happening Proximity codes: Safety culture (10) Separated rooms for rest and communication Proximity codes: - SVP interaction with different professions and disciplines Proximity codes: Roles and responsibilities (13) SVP occurrence and interactions with the responsibilities of different healthcare providers Proximity codes: Interprofessional care (13), safety culture (18), qualitative staff resources (12), disappointment (14), guilt (14), and rationalization (13) Predisposition depends on the safety culture level of an institution Proximity codes: Roles and responsibilities (18), rationalization (11), hot and cold debriefing (13), communication peers (14) SVP phenomena due to staff shortages Proximity codes: Roles and responsibilities (12) SVP phenomena due to educational gaps and misqualification Proximity codes: - National factors, laws, regulations, and others Proximity codes: Roles and responsibilities (7) Detection of (primarily linguistic) gender diversity aspects Proximity codes: Roles and responsibilities (21) Factors of predisposition by moral distress and moral injury Proximity codes: Roles and responsibilities (21) Death or life-threatening situation (e.g., shock, myocardial infarction, sepsis) Proximity codes: - Nutritional situation (e.g., starving, cachexia) Proximity codes: - Issues with the airway and airway devices (choking, tracheal cannula problems) Proximity codes: - Conflicts with other persons in a stressful situation Proximity codes: - Emotional binding to a patient experiencing a difficult situation Proximity codes: - Situations of feeling overwhelmed or helpless Proximity codes: - A near mistake without any harm Proximity codes: - A difficult, unexpected situation in routine care Proximity codes: - Feeling of pain Proximity codes: - Recurrent phases of remembering the situation Proximity codes: - Feeling disappointed by the situation Proximity codes: Roles and responsibilities (13), safety culture (7) Feeling interested in the situation or phenomenon, information seeking. Proximity codes: Roles and responsibilities (9), rationalization (13), hot and cold debriefing (14), communication with peers (23), thriving (13), “multiple” victim (18) Feeling ashamed of one’s own decisions or actions Proximity codes: - Feeling guilty for the adverse event Proximity codes: Roles and responsibilities (14), safety culture (10), restoring personal integrity (11), rationalization (9), thriving (10) Phrases expressing fear Proximity codes: Roles and responsibilities (13), safety culture (10), rationalization (9), knowledge and learning (7), thriving (10), intrusive reflections (9), restoring personal integrity (11) Phrases expressing anger Proximity codes: - Phrases expressing enjoyment Proximity codes: - Phrases expressing disgust Proximity codes: - Phrases expressing sadness Proximity codes: - Repression of the situation Proximity codes: - Justification of one’s own role in the event Proximity codes: - Trying to understand for oneself what and how it happened Proximity codes: Roles and responsibilities (13), safety culture (11), restoring personal integrity (17), interest (13), knowledge and learning (9), fear (9) Retrieving knowledge about the events and building competence Proximity codes: Restoring personal integrity (11), rationalization (9), passing time (8), peer communication (13), family support (11), thriving (21) Teaching others about the event to prevent them from making the same mistake Proximity codes: - Being mindful of oneself Proximity codes: - Being absent from the situation and taking rest Proximity codes: - Being absent to avoid situations or persons Proximity codes: - Being present to follow workplace structures and daily routines Proximity codes: - Physical activities Proximity codes: - “Time heals all wounds” Proximity codes: - Alcohol for relief after an event (NOT alcoholism) Proximity codes: - Coded for changes in an organization or institution after an event, preventing the event from occurring again, or reducing the risk Proximity codes: Restoring personal integrity (8), roles and responsibilities (6), patient safety (7), airway (6), thriving (10) Performing a direct (hot) debriefing after an event or a structural debriefing days later (cold) Proximity codes: Roles and responsibilities (10), patient safety (13), restoring personal integrity (16), emotional first aid (14), interest (14), rationalization (13), communication with peers (25), thriving (10) Communicating with the supervisor about the event Proximity codes: - Peer support from persons of another discipline or profession Proximity codes: Roles and responsibilities (7), interprofessional care (5), emotional first aid (6), professional therapy (7) Peer support without any further definition Proximity codes: Emotional first aid (16) Peer support from persons of the same discipline or profession Proximity codes: Roles and responsibilities (14), patient safety (14), restoring personal integrity (21), emotional first aid (33), thriving (14), interest (23), fear (8), rationalization (17), knowledge and learning (13), passing time (13), hot and cold debriefing (25), family support (12) Support from family members Proximity codes: Knowledge and learning (9), restoring personal integrity (8), emotional first aid (10), communication (peers) (12), thriving (8) Support from patients and proxies who were involved or harmed Proximity codes: - All support is provided by professionals who are educated to deal with second victims and traumatized persons Proximity codes: Roles and responsibilities (6), emotional first aid (6), “multiple” victim (10), drop-outs (2), survivors (4), unspecified outcome (9) Not a second victim Proximity codes: - Dysfunctional behavior, but staying in place Proximity codes: - Dropping out of the workplace after events Proximity codes: - Growing from the event Proximity codes: - Unknown outcome Proximity codes: - A person actively denying being a second victim multiple times Proximity codes: - Persons reporting one event Proximity codes: - Persons reporting more than one event Proximity codes: - Persons not identifying as a second victim but reporting events and traumatization according to the second victim definition Proximity codes: - Persons reporting that they had forgotten about their second victim experience and realized it during the survey Proximity codes: - The evidence column shows exemplary articles and manuscripts that described the “code” earlier. Codes with an asterisk (+) are “central” codes that show many relationships to other codes and subcodes in the code-browser-matrix. These proximity codes are labeled with a number, showing the number of the linked code in proximity to the described code. Codes with an * are those that could potentially inform new hypotheses For Scott’s model [12], the subsequent phases were coded as follows: “chaos and response phase” (n = 5), “intrusive reflections” (n = 7), “restoring personal integrity” (n = 15), “enduring the inquisition” (n = 5), “emotional first aid” (n = 32), “surviving” (n = 5), “growing” (n = 8), and “leaving” (n = 3). The coding process and derived themes are shown in Table 6. After checking the descriptions of the codes in prior studies (Table 6), we found the following new subcodes (outlined in detail in Table 6) that may potentially generate new hypotheses concerning second victim effects.“Nutrition” (2.1.2) as a cause of the second victim effect (e.g., by withdrawal or withholding oral intakte, but also malnutrution)“Enjoyment” (3.3.3) following a second victim effect (“injoyment” and positive emotion was mainly feedbacked related to our conducting this study, as SVP is thematized by the study)“Teaching” (4.1.2.2.) to cope with an SVP (e.g., that educative measures can prevent errors from the second victim to others)“Support from first victims” (4.5) (e.g., forgiving)“Silent victims” (5.2.4) (persons initially not identifying, but later acknowledging the situation)“Repression” and “recall bias” (4.1.1. and 5.2) (both indicating forgetting and coping as possible barriers to addressing the question of having experienced the phenomenon) “Nutrition” (2.1.2) as a cause of the second victim effect (e.g., by withdrawal or withholding oral intakte, but also malnutrution) “Nutrition” (2.1.2) as a cause of the second victim effect (e.g., by withdrawal or withholding oral intakte, but also malnutrution) “Enjoyment” (3.3.3) following a second victim effect (“injoyment” and positive emotion was mainly feedbacked related to our conducting this study, as SVP is thematized by the study) “Enjoyment” (3.3.3) following a second victim effect (“injoyment” and positive emotion was mainly feedbacked related to our conducting this study, as SVP is thematized by the study) “Teaching” (4.1.2.2.) to cope with an SVP (e.g., that educative measures can prevent errors from the second victim to others) “Teaching” (4.1.2.2.) to cope with an SVP (e.g., that educative measures can prevent errors from the second victim to others) “Support from first victims” (4.5) (e.g., forgiving) “Support from first victims” (4.5) (e.g., forgiving) “Silent victims” (5.2.4) (persons initially not identifying, but later acknowledging the situation) “Silent victims” (5.2.4) (persons initially not identifying, but later acknowledging the situation) “Repression” and “recall bias” (4.1.1. and 5.2) (both indicating forgetting and coping as possible barriers to addressing the question of having experienced the phenomenon) “Repression” and “recall bias” (4.1.1. and 5.2) (both indicating forgetting and coping as possible barriers to addressing the question of having experienced the phenomenon) Furthermore, by quantitatively exploring the code browser matrix, we found the code “roles and responsibilities” (1.1.4) to be most prominent, with multiple instances of coding in the vicinity of other codes. All phrases and codes thus derived are listed in the Supplementary Materials. Discussion This study explored the presence of second victim effects among healthcare providers treating patients with dysphagia in German-speaking countries. Further, it explored the detectability of the effects to investigate persons not identified as second victims, but suffering from related effects. We observed that SVP is detectable among German-speaking healthcare providers in their professional efforts to support patients with dysphagia. We learned that “silent victims”, individuals who do not identify with the term but later report or recall incidents that have caused them suffering, were not accurately detectable. Our findings emphasize that the detectability of the phenomenon must be re-evaluated, as current screening tools may underestimate the incidence of SVP with respect for sensitivity. Consequently, future research should concentrate n the refinement of screening instruments or the development of new instruments to detect “silent victims” too. In explorative regression analysis, we detected a possible but not significant association of self-identifying a second victim with neuroticism, a risk factor identified in prior studies [19, 20]. Our results align with those of preceding SeViD studies, which show a comparable prevalence of SVP and similar causes (unexpected death, patient harm, and near-miss cases) among young physicians, nurses, emergency medicine physicians, pediatricians, general practitioners, and family caregivers [16–22]. Allender et al. also reported a high rate of SVP among respiratory therapists in the USA [28], which is consistent with our results. Furthermore, our findings regarding the risks associated with airway management are well-documented in the literature, such as difficult airways and medication errors reported in the fields of anesthesiology and critical care [46]. The regression analysis also yielded results similar to those of other SeViD studies, with neuroticism possibly being a predictor of reporting the experience of SVP. However, this finding must be interpreted with caution, due to model performance, non-confirming Bootstrap confidence intervals, underrecognition of a low “event per variable” ratio (1:4.7), and possibly limited power concerning a classical ratio of at least 1:10 [58] and a more liberal ratio of 1:5 [59]. Our findings regarding the well-being of healthcare providers treating dysphagic patients, who play a crucial role in multiple healthcare domains and intensive care, are concerning. This is especially important in light of a significant rate of persons reporting being affected for more than one year and without recovery, the recent and anticipated staff shortages [45], the rising prevalence of dysphagia and an aging and multimorbid population [2, 5, 45]. One core finding is the high rate of “silent victims,” indicating underestimation or a misclassification bias of the screening question toward the negative state. This may be a result of repression or recall bias, as seen in the qualitative analysis, or label avoidance, social desirability, or the preservation of personal integrity. Future studies should aim to better detect silent second victims, for example, by using alternative tests that reduce self-assessment. Options include evaluating other instruments like interviews or second-person rating (as an example: the Hamilton Score for depression, first introduced in 1960 [60] that bases on the assessment of an interview and is still in use in its 17th version). As self-assessment is integrated in many psychological questionaires, further development of the SVEST and SeViD questionnaires is important, for example by implementation of indirect questions. However, these methods face several limitations on their own [61]. Unfortunately, we recruited too few participants to complete the GSVEST-R or MISS-HP scale, we were unable to clarify the role of these instruments in our analysis. Our findings suggest opportunities for future research on silent victims, which could include single- or double-item screening questions (e.g., as a first screening step), subjective symptom assessments (as a second step), and multi-item assessments (as a third step). However, hidden or concealed experiences indicate that professional interviews or longitudinal instruments are likely more helpful than cross-sectional questionnaires. The findings regarding silent victims also shed light on the parameter of neuroticism that was not significant in our regression analysis and contradicting prior findings [17, 19, 20]. This is possibly a result of insufficient power with need of larger sample sizes for confirmation. Considering it to be a factor, the combination with silent victims may hypothetically reframe the meaning of neuroticism: Assuming that nearly every healthcare provider is experiencing a second victim effect, as suggested by Von Laue [15] and our findings, neuroticism may not only be a risk factor but may also lead to more frequent or prolonged reporting. Hypothetically, the trait may be reframed from a risk factor to a protective factor, as reporting about such effects provides the option to act on related experiences. However, additional research is needed to confirm this speculative hypothesis, as our results are exploratory and require confirmatory trials with larger sample sizes. The qualitative results are consistent with previous findings, such as Scott’s results defining several phases of second victim experiences [12], our findings regarding the emotional impact of dysphagia using the Ekman model of emotion [48, 62], and with findings that led to the development of the SVEST and SVEST-R dimensions [30, 49]. However, we identified some aspects in the qualitative component of our survey that may warrant reconsideration in more in-depth qualitative research, involving methods such as focus groups or interviews, which could help refine the instruments. Specifically, codes for “self-support,” “support from first victims,” and “support from family members” were identified. Interestingly, the last factor was part of the original 10-factor SVEST-R but was excluded during the validation process, resulting in a 9-factor instrument [30]. The current results strongly suggest the need for educational and support programs for health care providers involved in dysphagia therapy. This is indeed a necessary step, and various programs have been implemented thus far [57, 63–66]. Accordingly, the following recommendations are made: Education on dysphagia and related conditions like tracheostomy and managing adverse events and emotional impact is needed to prevent situations leading to SVP (e.g., airway difficulties, management conflicts arising from or leading to errors [67], and emotional problems associated with dysphagia [62]).Second, resilience and victim support programs as well as crew management methods (“CRM”) are needed to prevent and manage these situations in routine care and within teams for example by peer and supervisor support [10, 14, 63, 64, 66, 68], structured prebriefings, hot and cold debriefings [69] other instruments like balint groups or morbidity and mortality conferences. Education on dysphagia and related conditions like tracheostomy and managing adverse events and emotional impact is needed to prevent situations leading to SVP (e.g., airway difficulties, management conflicts arising from or leading to errors [67], and emotional problems associated with dysphagia [62]). Education on dysphagia and related conditions like tracheostomy and managing adverse events and emotional impact is needed to prevent situations leading to SVP (e.g., airway difficulties, management conflicts arising from or leading to errors [67], and emotional problems associated with dysphagia [62]). Second, resilience and victim support programs as well as crew management methods (“CRM”) are needed to prevent and manage these situations in routine care and within teams for example by peer and supervisor support [10, 14, 63, 64, 66, 68], structured prebriefings, hot and cold debriefings [69] other instruments like balint groups or morbidity and mortality conferences. Second, resilience and victim support programs as well as crew management methods (“CRM”) are needed to prevent and manage these situations in routine care and within teams for example by peer and supervisor support [10, 14, 63, 64, 66, 68], structured prebriefings, hot and cold debriefings [69] other instruments like balint groups or morbidity and mortality conferences. Both actions of primary and secondary prevention are beneficial and justifiable from the individual perspective, when regarding the potential economic impact of a “lost” second victim, as well as from institutional and national perspectives [70, 71]. Further, relatively new in the context of SVP (but not dysphagia) are aspects related to oral intake and social desirability. SVP may be provoked by compassion or conflict, for example, when a specialist is involved in relevant medical decisions (e.g. gastrostomy). Health care providers treating patients with dysphagia may be at risk of SVP when patients or family members do not accept dysphagia and express a desire to eat “normally”, while leaving the subjective or objective responsibility to the therapist. Possible conflicts in medical diagnose and guideline recommendations, ethical and legal regulations, own issues of moral, ethics and beliefs (“moral compass”), committment to the patient’s more or less known will and respect to caring and burdened family members is challenging and may impact psychological well-being. Therefore, the central code “roles and responsibilities” is of high interest, as many participants stated that this is a significant problem in facing serious situations and events with need for clarification in situations (e.g. an acute event like airway obstruction with desaturation) and more generally in institutions (“who is responsible for the per oral nutrition in tracheostomized patients on ward xy?”). Consequently, this aspect of second victim experiences should be investigated further, especially because roles and responsibilities play a substantial role in resuscitation settings and clarity on the won roleis known to optimize team communication and crew resource management [41]. Additionally, our findings on coping mechanisms require clarification, for example, for those providers using teaching as a coping mechanism as mentioned by one participant, who uses his experience to avoid that his/her students make the same mistake. It is still unclear to what extent a second victim effect might influence the performance of practically active or inactive medical teachers and educators, as well as the competence gain of their students, leaving a wide field of investigation. This is particularly relevant assuming that nearly every medical profession experiences second victim effects and that every educator is personally affected. Limitations Our explorative study has the following limitations. First, convenience-based surveys are prone to selection bias and other problems related to cross-sectional questionnaires, including a high survey burden, survey fraud [72], and false clicks (which may have occurred when two second victims denied an event afterwards; see Table 4). Selection bias may have resulted in higher prevalence rates in the sample, especially because only those interested in the topic were likely to answer. Further, we did not include those who have already left the job (“survivorship bias”). However, even if the rate hypothetically would be lower (e.g. “only” 10%), this would affect a significant proportion of individuals involved in dysphagia and respiratory therapy, with every loss being unacceptable for patients, teams, institutions, and the system. Further, details regarding the number of health care providers specializing in dysphagia and the number of emails forwarded from the heads of departments to their co-workers could not be obtained. Despite these limitations, the phenomenon is detectable, comparable to larger studies in other populations, and may be critical particularly in cases of absenteeism, dysfunctional behavior, and defensive medicine for dependent patients within a complex healthcare system. Finally, the qualitative analysis included only free-text entries, necessitating more in-depth evaluations through interviews, focus groups, and longitudinal designs to overcome the common limitations of cross-sectional and convenience sampling methods. Conclusion SVP is detectable in healthcare provides involved in dysphagia recognition and treatment. Difficult airway issues, unexpected events, conflicts, and feeding-related aspects are typical triggers of long-lasting symptoms. Furthermore, a significant proportion of those who denied being second victims were actually affected. This raises the question about the prevalence and experiences of “silent victims” and the capacity of developed instruments to detect them. With silent victims, new questions arise concerning the quality of existing instruments, the overall course of the second victim phenomenon, inter- and intraindividual differences, and the needs and properties of primary and secondary prevention and rehabilitation. Concerning risk factors, we found that neuroticism may be not only be a risk factor for experiencing the phenomenon but also a protective one. Additionally, our qualitative analysis revealed several new aspects of the second victim experience, such as triggers (oral intake), as well as other factors such as support from family members and first victims and the option to report about the situation and educate others. Overall, several questions have emerged from this pilot study, indicating the need for deeper longitudinal research and exploration to design efficient prevention and support programs to address SVP and related experiences among healthcare professionals. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2