Prevalence, components and consequences of moral injury: preliminary validation of a new brief measure

doi:10.21203/rs.3.rs-2222013/v1

Download PDF

Short Report

Prevalence, components and consequences of moral injury: preliminary validation of a new brief measure

https://doi.org/10.21203/rs.3.rs-2222013/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this older preprint version

Read the latest preprint version →

Aim

The aim of this study was to design and use a parsimonious survey tool to use in real time to assess moral injury and describe how moral injury relates to burnout and intent to leave the job. The Moral Injury Quotient (MIQ) is derived from this 6-item tool.

Findings

The MIQ metric has good performance characteristics and captures a substantiative portion of moral injury. It is related to clinically meaningful changes; each 10-point increase in the MIQ was associated with 125% increased odds of burnout (p < 0.001) and 50% increased odds of intending to leave the job (p < 0.001).

Conclusions

Measuring and addressing moral injury has major workforce implications. This metric is an “off the shelf” tool that may be helpful in busy clinical settings to assess, implement improvements, and reassess for reductions in moral injury.

Moral distress

Moral injury

Burnout

turnover

Workforce shortages have accelerated during COVID 19 and are contributing to challenges with retention, recruitment, staffing, and the capacity of healthcare systems to meet patient needs. Sinsky found that 40% of nurses and 24% of physicians intend to leave their jobs (1). A Washington Post poll revealed 80% of physicians were exploring other jobs (2). Losses of $4.6 billion annually are attributed to burnout-related physician turnover (3), and hospitals can lose $5 million annually if not attending to nurse wellbeing and burnout (4). These health professional shortages will continue to worsen unless system-level strategies are implemented to support healthcare workers (HCWs) and mitigate burnout and related challenges.

While recent research has examined burnout among healthcare professionals (5), there is less focus on moral distress or injury. Moral distress occurs when someone has an ethical construct of the best action but is unable to take that action; Moral injury (MI) occurs when moral distress is not addressed and becomes repeated or chronic (6). Moral injury appears to be common in the healthcare workforce (7, 8), particularly in nursing (7). Because MI may have lasting consequences on the workforce and populations being served, this gap in knowledge about it requires attention. Common examples of contributors to moral distress include providing non-beneficial care due to family insistence; participating in suboptimal patient care due to lack of resources; and receiving insufficient administrative support to address problems thought to be compromising patient care. The COVID era has presented everyone in health care, particularly intensive care unit (ICU) teams, with amplified versions of MI which, if not addressed, can contribute to physical and emotional trauma and sequelae, ineffective or unsafe care delivery, and for some, a decision to leave healthcare altogether.

To our knowledge, there are few if any clinically employed metrics for MI. Without a way to measure MI we risk continuing to acknowledge but not understand the scope of the problem and to lack data to drive interventions and system change. Thus, there could be value in a brief, valid MI measure to track progress to eliminating it in HCWs. We adapted items from Epstein and colleagues’ Measure of Moral Distress for Healthcare Professionals (MMD-HP, 9) and used focus groups to develop a brief instrument. Here we report performance characteristics of the measure, and relationships of MI to malleable individual and health system characteristics, such as reflection time after challenging encounters and availability of ethics support. Finally, we link these characteristics to burnout and intent to leave, using the words of nurses and other HCWs in our ICUs.

A brief MI inventory was created in part, from MMD-HP (9) items, and from focus groups with critical care physicians, advanced practice clinicians (APCs) and nurses in May 2021. Once the instrument was developed, 319 critical care physicians, APCs and nurses at Hennepin Healthcare were surveyed in June 2021. The Hennepin Healthcare IRB determined the study to be a program evaluation (IRB review declined 4/14/21). The gold standard MMD-HP has 27 items, each rated for frequency and level of distress. Our focus groups identified five prevalent morally distressing scenarios from Epstein’s 27; another identified theme was distress from care complicated by health care disparities. These 6 domains comprised the MI instrument. The product of “distress (scored 0–4) X frequency (scored 0–4)” contributed to a “composite score” (range 0–16) for each domain. Composite scores were summed for a Moral Injury Quotient (MIQ instrument, Appendix 1), scored from 0–96 (6 X 16). Cronbach’s alpha and an intercorrelation matrix assessed internal consistency. Logistic regression models assessed relationships of MI to binary outcomes, adjusting for race, gender and role. Two-sided p-values determined statistical significance at the 0.05 level (R Core Team, 2021, version 4.1.1). Our 6 components included lack of resources, lack of administrative support, non-beneficial care, abusive families, care not alleviating suffering, and care complicated by inequities.

Additional questions provided contextual insights and correlational validity for the MIQ; these included two open-ended questions, the validated single item burnout question (10), and two items assessing likelihood of leaving in 2 years or reducing hours in one year. (“Moderately”, “likely”, and “definite” meant intent to leave (ITL), and “maybe, likely, and definitely” meant intent to reduce (ITR) hours). MIQ with scoring is in the Appendix.

Survey metrics. The MIQ sum of composite scores had a mean value of 48 (Fig. 1). Twenty five percent of MIQs were > 60 (Fig. 1), and 32% of respondents indicated 4 of 6 moral injury components were “highly distressing and/or very frequent.” Cronbach’s alpha was 0.71 for distress, 0.75 for frequency and 0.83 combined (> 0.7 = good to very good internal consistency). In the intercorrelation matrix, 48% of items correlated at r > 0.3.

Mean MIQ for nurses was 50.3 vs 32.8 for physicians and APCs (higher more distressing, p < 0.0001). Composite (0–16) domain scores were higher for nurses in lacking resources (10.5 vs 5.3 in physicians and APCs, p < 0.001), lack of administrative support (9.3 vs 6.2, p < 0.02) and abuse by patients/families (8.2 vs 4.2, p < 0.001). Adjusted logistic regression models showed clinically meaningful Regression models showed with increases in MIQ, the adjusted odds of burnout increased significantly (Odds Ratio (OR) for burnout with 10-point MIQ increase 2.257 (95% CIs 1.622, 3.361), p < 0.001), as did odds of intending to leave (OR of intent to leave with 10-point MIQ increase 1.504, (95% CIs 1.187, 1.966), p < 0.001). Construct validity was supported by clear questions derived with expert input, satisfactory internal consistency, MIQ correlation with the valid burnout measure (11), and correlation with findings from open-ended comments (see below).

Survey results. Of 319 individuals surveyed, 129 (40%) responded (84% were nurses, 14% physicians or APCs, 70% females and 83% white). Burnout was reported by 67% (72% of nurses vs 40% of physicians and APCs, p = 0.01). Lacking resources was the most frequently reported reason for distress. This was followed in composite score rank by lack of administrative support, non-beneficial care, working with abusive families/patients, seeing unaddressed structural inequities, and care not alleviating suffering. Intent to leave was noted by 52%, and ITR by 63%; rates of ITR and ITL did not differ statistically between physicians/APCs and nurses.

Open-ended survey comments gave voice to numerous concerns, particularly among nurse respondents. Nurses’ responses to the question soliciting situations that were morally distressing aligned with the six MIQ domains (Table 1). Nurses also identified other sources of MI, including disruptions to teamwork, fewer opportunities for education or skill-building, lack of professionalism, and visitor policies. Suggested solutions included increased staffing, psychological support, and demonstrable support from administrative leaders. Comments that could support work environment optimization included those related to control of nurse schedules and delivery of non-beneficial patient care. Open-ended comments generally supported the validity of the MIQ measure in identifying important areas of moral concern.

Table 1

Representative quotes regarding morally distressing situations from nurse perspective
MIQ Domain	Quote
Lack of resources	Sometimes I feel like I’m a bad nurse and not providing all the care I should because we don’t have enough staff to help after the recent staffing cuts.
Lack of administrative support	Working with the same administrators who tried to minimize our safety concerns due to a lack of adequate Personal Protective Equipment (PPE) during the initial phases of COVID. It's very distressing that I have to take direction from those people.
Non-beneficial care	When the family overturns an advanced directive a patient made so we can do painful and futile care
Abusive families	Family conflict and verbal abuse from visitors (at least every other day)
Care not alleviating suffering	The quality of life for a patient is no longer there yet family/friends continue with care even though they are not here to see first-hand the stress they put the patient through.
Care complicated by inequities	…patients who are just sent back to difficult living conditions with no resources to improve their lives, the vast limitations and inequities in the US healthcare system.

The new MIQ metric, with 6 items and good performance characteristics, appears to capture a substantive portion of injury and is related to clinically meaningful changes in burnout and ITL. These findings could have important implications for the current and future healthcare workforce.

Among 129 ICU nurses and clinicians, we found worrisome rates of burnout and MI correlated with reports of intent to leave or reduce hours. Significantly more nurses reported moral distress, injury, and burnout as compared with those in other roles. ICU practice is full of ethical decisions and some moral distress is expected. Epstein and Hamric describe a crescendo effect of moral distress with unresolved residue that worsens with each episode of moral distress (12). This residue is best addressed at the system level but individual-level strategies cannot be overlooked. Enhancing resources for patient care, communication and workflow streamlining, as well as ethical support and policies to address requests for non-beneficial treatment are examples of system-level intervention. Research on mitigating nurse and clinician burnout demonstrates promising outcomes from self-care strategies (e.g., physical activity, mindfulness, boundary setting), but these may be insufficient if not supported by system-level changes (13). Open ended comments from our ICU physicians, APCs and nurses offer heartfelt statements of challenge, as well as actionable suggestions for means to reduce burnout and MI.

Our study has limitations, with a cross-sectional design involving only one metropolitan hospital. A brief survey may miss the full spectrum of contributors to MI (8, 9); further, the instrument while correlated with other relevant items, was not compared to other existing tools. Our next steps include using this tool with a larger battery of instruments assessing MI, burnout, ITR/ITL, and to test interventions to eliminate MI in varying settings.

Many nurses and clinicians are not experiencing the level of well-being that HCW’s deserve, with negative downstream consequences affecting their families, colleagues, and patients. COVID-19 has exaggerated and unearthed complex cascading factors that must be understood and intervened upon so the workforce responsible for caring for and healing others is themselves, cared for and healthy.

HCW – Healthcare Worker

MI – Moral Injury

ICU – Intensive Care Unit

MMD-HP – Measure of Moral Distress for Healthcare Professionals

APC – Advanced Practice Clinicians

MIQ – Moral Injury Quotient

ITL – Intent To Leave

ITR – Intent To Reduce

PPE – Personal Protective Equipment

Ethics approval and consent to participate The Hennepin Healthcare IRB determined the study to be a program evaluation project (IRB review declined 4/14/21). Ethics approval was also not deemed necessary as this was a program evaluation project.

Consent for publication. Does not apply as all data is anonymous, even to the investigators

Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests: Dr. LeClaire was supported in part through her employer at the time the study was conducted (Hennepin Healthcare) by the AMA for burnout reduction studies. Dr. Linzer is supported through his employer, Hennepin Healthcare, for burnout reduction studies by AMA, IHI, ACP, ABIM, Optum Office for Provider Advancement (OPA), Essentia Health System, Gillette Children’s Hospital, St Paul MN, and the California AHEC System. He performs scholarly work for the NIH and AHRQ and consults for Harvard University on a grant linking work conditions to diagnostic accuracy (funds donated to Hennepin Healthcare Foundation). Ms. Poplau is supported by AMA, IHI, ACP, Essentia, Optum, Gillette and California AHEC. Ms. Freese is supported by University of Minnesota CTSI (see acknowledgement). Dr. Porta is supported through her employer, University of Minnesota.

Funding: This research was supported by the National Institutes of Health’s National Center for Advancing Translational Sciences, grant UL1TR002494

Authors' contributions:

MLC- primary investigator and ran the focus groups, constructed the survey, dispersed the data in the organization, and prepared brief report

SP- The data collector through survey

RLF- Statistician involved in study design and data analysis

CP- Substantial report revisions, including nursing expertise, added important references

EES- Qualitative analysis of open-ended questions

ML- Co-investigator with Dr. LeClaire (MLC). Worked on study design, supported survey, and prepared this report in conjunction with MLC

All authors have viewed the final version of this Brief Report

Acknowledgements: This research was supported by the National Institutes of Health’s National Center for Advancing Translational Sciences, grant UL1TR002494. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health’s National Center for Advancing Translational Sciences. The authors also acknowledge Elizabeth Epstein and the late Ann Hamric for their pioneering work on moral distress.

Sinsky CA, Brown R, Linzer M, et al. COVID-related stress, burnout, and work intentions of US healthcare workers. Mayo Clin Proc Inn Qual Out. 2021;5(6):1165-73.
Kaiser Family Foundation-Washington Post. Frontline Health Care Workers Survey. March 2021. https://context-cdn.washingtonpost.com/notes/prod/default/documents/4d8d1ddf-c192-40f9-9e3a-7a3fefa0d928/note/91e5f1ac-2cc5-41bb-b164-ecb4d77ed0b5.#page=1. Accessed 28 October, 2022.
Han S, Shanafelt T, Sinsky C, et al. Estimating the attributable costs of physician burnout in the United States. Ann Int Med. 2019;170(11);784-90.
Muir K, Wanchek T, Lobo J, et. al. Evaluating the costs of nurse burnout- attributed turnover: a Markov modeling approach. J Patient Safety. 2022;18(4):351-57.
Prasad K, McLoughlin C, Stillman M, et al. Prevalence and correlated of stress and burnout among US healthcare workers during the COVID-19 pandemic. Eclinical Med. 2021;35:100879.
Dean W, Talbot S, Caplan A. Clarifying the language of clinician distress. JAMA. 2020;323(10):923-24.
LeClaire M, Poplau S, Linzer M, et al. Compromised integrity, burnout and intent to leave the job in critical care nurses and physicians. Crit Care Explor. 2022;4(2):e0629. doi: 10.1097/CCE.000000000000062
Dodek PM, Cheung EO, Burns KEA, et al. Moral distress and other wellness measures in Canadian Critical Care physicians. Ann Am Thorac Soc. 2021;18(8):1343-51. doi:10.1513/AnnalsATS.202009-1118OC
Epstein E, Whitehead P, Hamric A, et al. Enhancing understanding of moral distress: the Measure of Moral Distress for Health Care Professionals. AJOB Emp Bioethics. 2019;10(2):113-24.
Olson K, Sinsky C, Linzer et al. Cross-sectional survey of workplace stressors associated with physician burnout measured by the Mini-Z and the Maslach Burnout Inventory. Stress and Health. 2018;35(2):157-75.
Cook D, Beckman T. Current concepts in validity and reliability for psychometric instruments: theory and application. Am J Med. 2006;119(2):166.e7-16.
Epstein E, Hamric A. Moral distress, moral residue, and the crescendo effect. J Clin Ethics. 2009:20(4):330-42.
Dulko D, Zangaro G. Comparison of factors associated with physician and nurse burnout. Nurs Clin N Am. 2022;57(1):53-66. https://doi.org/10.1016/j.cnur.2021.11.004.

No competing interests reported.

AppendixMoralInjuryQuotient103122.pdf

Download PDF

Version 1

posted

You are reading this older preprint version

Read the latest preprint version →

Prevalence, components and consequences of moral injury: preliminary validation of a new brief measure

Status:

Version 1

Abstract

Figures

Introduction:

Methods:

Results:

Discussion:

List Of Abbreviations:

Declarations:

References:

Additional Declarations

Supplementary Files

Status:

Version 1