Advertisement

Patient-Directed Discharges Among Persons Who Use Drugs Hospitalized with Invasive Staphylococcus aureus Infections: Opportunities for Improvement

Open AccessPublished:September 08, 2021DOI:https://doi.org/10.1016/j.amjmed.2021.08.007

      Abstract

      Background

      Despite the high burden of Staphylococcus aureus infections among persons who use drugs, limited data exist comparing outcomes of patient-directed discharge (known as discharge against medical advice) compared with standard discharge among persons who use drugs hospitalized with S. aureus infection.

      Methods

      We conducted a retrospective study of hospitalizations among adults with S. aureus bacteremia, endocarditis, epidural abscess, or vertebral osteomyelitis at 2 San Francisco hospitals between 2013 and 2018. We compared odds of 1-year readmission for infection persistence or recurrence and 1-year mortality via multivariable logistic regression models adjusting for age, sex, Charlson comorbidity index, and homelessness.

      Results

      Overall, 80 of 340 (24%) of hospitalizations for invasive S. aureus infections among persons who use drugs involved patient-directed discharge. More than half of patient-directed discharges 41 of 80 (51%) required readmission for persistent or recurrent S. aureus infection compared with 54 of 260 (21%) patients without patient-directed discharge (adjusted odds ratio 3.8, 95% confidence interval [CI] 2.2-6.7). One-year cumulative mortality was 15% after patient-directed discharge compared with 11% after standard discharge (P = .02); however, this difference was not significant after adjustment for mortality risk factors. More than half of deaths in the patient-directed discharge group (7 of 12, 58%) were due to drug overdose; none was due to S. aureus infection.

      Conclusions

      Among persons who use drugs hospitalized with invasive S. aureus infection, odds of hospital readmission for infection were almost 4-fold higher following patient-directed discharge compared with standard discharge. All-cause 1-year mortality was similarly high in both groups, and drug overdose was a common cause of death in patient-directed discharge group.

      Keywords

      Clinical Significance
      • Approximately 1 in 4 hospitalizations for invasive Staphylococcus aureus infections among persons who use drugs involved a patient-directed discharge, which is known as discharge against medical advice.
      • Adjusted odds of hospital readmission for infection was 3.8 times higher after patient-directed compared with a standard discharge.
      • Adjusted odds of 1-year, all-cause mortality were similarly high in patient-directed and standard discharge groups. In the patient-directed discharge group, deaths were primarily due to drug overdose.

      Introduction

      Invasive Staphylococcus aureus infections associated with drug use are increasingly common causes of hospitalization.
      • Ronan MV
      • Herzig SJ.
      Hospitalizations related to opioid abuse/dependence and associated serious infections from 2002 to 2012.
      • Schranz AJ
      • Fleischauer A
      • Chu VH
      • Wu L-T
      • Rosen DL.
      Trends in drug use–associated infective endocarditis and heart valve surgery, 2007 to 2017.
      • McCarthy NL
      • Baggs J
      • See I
      • et al.
      Bacterial infections associated with substance use disorders, large cohort of united states hospitals, 2012–2017.
      • Jackson KA
      • Bohm MK
      • Brooks JT
      • et al.
      Invasive methicillin-resistant Staphylococcus aureus infections among persons who inject drugs — six sites, 2005–2016.
      Persons who use drugs hospitalized with invasive S. aureus infections also face elevated and increasing rates of patient-directed discharge or discharge against medical advice.
      • Suzuki J
      • Robinson D
      • Mosquera M
      • et al.
      Impact of medications for opioid use disorder on discharge against medical advice among people who inject drugs hospitalized for infective endocarditis.
      • Ti L
      • Ti L.
      Leaving the hospital against medical advice among people who use illicit drugs: a systematic review.
      • Kim J-H
      • Fine DR
      • Li L
      • et al.
      Disparities in United States hospitalizations for serious infections in patients with and without opioid use disorder: a nationwide observational study.
      A recent study found that odds of patient-directed discharge during hospitalization for drug use-related endocarditis have been increasing 12% per year nationwide.

      Kimmel SD, Kim J-H, Kalesan B, Samet JH, Walley AY, Larochelle MR. Against medical advice discharges in injection and non-injection drug use-associated infective endocarditis: a nationwide cohort study [e-pub ahead of print]. Clin Infect Dis. doi: 10.1093/cid/ciaa1126. Accessed March 3, 2021.

      These interrupted acute care episodes are associated with patient dissatisfaction and distrust of health care institutions,
      • Alfandre D
      • Schumann JH.
      What is wrong with discharges against medical advice (and how to fix them).
      elevated 30-day hospital readmission rates,
      • Choi M
      • Kim H
      • Qian H
      • Palepu A.
      Readmission rates of patients discharged against medical advice: a matched cohort study.
      • Glasgow JM
      • Vaughn-Sarrazin M
      • Kaboli PJ.
      Leaving against medical advice (ama): risk of 30-day mortality and hospital readmission.
      • Kumar N.
      Burden of 30-day readmissions associated with discharge against medical advice among inpatients in the United States.
      • Southern WN
      • Nahvi S
      • Arnsten JH.
      Increased risk of mortality and readmission among patients discharged against medical advice.
      • Yong TY
      • Fok JS
      • Hakendorf P
      • Ben-Tovim D
      • Thompson CH
      • Li JY.
      Characteristics and outcomes of discharges against medical advice among hospitalised patients.
      and in some studies, higher 30-day mortality.
      • Glasgow JM
      • Vaughn-Sarrazin M
      • Kaboli PJ.
      Leaving against medical advice (ama): risk of 30-day mortality and hospital readmission.
      ,
      • Southern WN
      • Nahvi S
      • Arnsten JH.
      Increased risk of mortality and readmission among patients discharged against medical advice.
      Despite the rising numbers of persons who use drugs with invasive S. aureus infections and rising patient-directed discharges, limited data compare infection outcomes, including mortality of persons who use drugs with invasive S. aureus infections that end in patient-directed discharge compared with standard discharge. Thus, we compare demographics, substance use characteristics, hospital readmission rates, and mortality rates between persons who use drugs with and without patient-directed discharge hospitalized with invasive S. aureus infection.

      Methods

      Study Design, Setting, and Participants

      We performed a retrospective cohort study of patients hospitalized from 2013 to 2018 at 2 hospitals, San Francisco General Hospital (226-bed public county hospital and trauma center) and UCSF Medical Center (839-bed tertiary care and referral hospital).
      We previously described the patient cohort;

      Appa A, Adamo M, Le S, et al. Comparative one-year outcomes of invasive Staphylococcus aureus infections among persons with and without drug use: an observational cohort study [e-pub ahead of print]. Clin Infect Dis. doi: https://doi.org/10.1093/cid/ciab367, accessed June 2, 2021.

      briefly, we identified 1,011 hospitalizations among adults during which a microbiologic diagnosis of invasive S. aureus infection (defined as bacteremia, endocarditis, epidural abscess, or vertebral osteomyelitis) was made via strutured chart review. For this analysis, we included only persons who use drugs and excluded hospitalizations that did not involve an invasive S. aureus infection or where S. aureus was designated a contaminant. In the present analysis of post-hospital discharge outcomes, we also excluded hospitalizations in which patients died or transitioned to comfort-focused care in the hospital. Our final sample consisted of 340 adult hospitalizations for invasive S. aureus infection.

      Definitions

      Four coauthors (AA, MA, SL, JD) performed structured chart reviews. We jointly reviewed a subset of charts to assure precision in data extraction. We defined persons who use drugs as patients with any non-prescribed substance use noted in hospital records of the primary team, infectious diseases consultants, or social workers. Cannabis, alcohol, and tobacco use were excluded. We defined patient-directed discharge as occurring against medical advice (or with synonymous terms, e.g. AMA or AWOL) in hospital discharge records or other documentation. Using electronic medical record (EMR) data, including data obtained from CareEverywhere (Epic Systems), we ascertained 30-day and 1-year readmission to any area hospital for ongoing or recurrent S. aureus infection. One-year all-cause mortality was ascertained both by chart review and cross comparison to the California state death registry.

      California Department of Public Health, Center for Health Statistics. California Electronic Death Registration. Available at: https://ca.edrs.us/edrs/secureLogin. Accessed June 24, 2021.

      We reviewed causes of death occurring <1 year after hospital discharge. Out-of-state patient deaths were not captured. We recorded patient-directed discharge reasons described in hospital records and identified common themes.

      Statistical Analysis

      We compared demographic, clinical, and substance use features between persons who use drugs with and without patient-directed discharge using χ2 tests. Adjusted odds of 1-year readmission for infection recurrence or persistence, as well as 1-year all-cause mortality, were compared between groups using a multivariate logistic regression model adjusted for age, race, housing status, medical comorbidities, and methicillin-resistant and methicillin-sensitive S. aureus (MRSA vs MSSA) infection.

      Ethics Statement

      The study was approved by the University of California, San Francisco Committee on Human Subjects Research.

      Results

      Overall, 80 of 340 (24%) hospitalizations for invasive S. aureus infection among persons who use drugs ended in a patient-directed discharge. Compared to persons who use drugs with a standard discharge, persons who use drugs with patient-directed discharge were younger (median age 44 vs 52 years), more likely to be experiencing homelessness (59% vs 31%), less chronically ill (median Charlson comorbidity score 1 vs 2), and more likely to be living with HIV (26% vs 17%; Table 1). Substance use characteristics also differed between patients with patient-directed discharge and standard discharges. Compared to patients who were discharged in a standard fashion, patients with self-directed discharge were more likely to report injection drug use (83% vs 67%), opioid use (69% vs 54%), methamphetamine use (61% vs 47%), and combined opioid and stimulant use (see Table 1).
      Table 1Demographic, Clinical, Substance Use, and Hospitalization Characteristics Among Persons Who Use Drugs Stratified by Discharge Status
      Patient-Directed Discharge
      Also known as against medical advice.
      N = 80
      Standard Discharge N = 260P Value
      P values were calculated using χ2 tests unless otherwise indicated.
      Demographic
      Age (median, IQR)44 (33-52)52 (42-57)<.01
      P value calculated using Wilcoxon rank sum test.
      Male sex (n, %)50 (63%)189 (73%).08
      Race/ethnicity.27
       White51 (64%)149 (57%)
       Black/African American13 (16%)57 (22%)
       Hispanic/Latinx6 (8%)35 (14%)
       Asian/Pacific Islander3 (4%)5 (2%)
       Other7 (9%)14 (5%)
      Experiencing homelessness (%, n)47 (59%)81 (31%)<.01
      Clinical
      Charlson comorbidity score (median, IQR)1 (0-2)2 (1-4)<.01
      P value calculated using Wilcoxon rank sum test.
      HIV positive21 (26%)43 (17%).03
      Immunosuppressed
      Immunosuppressed defined as those with HIV, chronic exposure to daily prednisone 10 mg or more, biologic agents or DMARDs, hematologic malignancy, or chemotherapy.
      (%, n)
      22 (28%)51 (20%).13
      Any mental health condition (%, n)30 (38%)89 (34%).59
      Substance Use
      Injection route described (%, n)66 (83%)176 (68%).04
      Recent drug use (<1 mo) (%, n)75 (94%)170 (65%)<.01
      Drug type (%, n)
       Heroin51 (64%)132 (51%).04
       Any opioid55 (69%)141 (54%).02
       Cocaine24 (30%)101 (39%)0.15
       Methamphetamine49 (61%)123 (47%).03
       Opioid and stimulant39 (49%)93 (36%).04
      Unhealthy alcohol use
      Diagnosed with alcohol use disorder or noted to consume more than 3 or 4 alcoholic drinks daily in women and men, respectively.
      (%, n)
      10 (13%)54 (21%).20
      Hospitalization
      MRSA infection46 (58%)119 (46%).07
      Type of infection (%, n)
       Bacteremia60 (75%)197 (76%).89
       Endocarditis17 (21%)47 (18%).62
       Vertebral osteomyelitis19 (24%)80 (31%).23
       Epidural abscess15 (19%)74 (28%).08
      ICU admission during hospital course14 (18%)105 (40%)<.01
      Surgical source control performed33 (41%)145 (56%).02
      PICC placed
      Of people eligible for PICC placement (ie, excluding people with preexisting central access).
      38/75 (51%)164/232 (71%)<.01
      Inpatient MOUD received
      Of those who reported opioid use.
      .51
       None26 (47%)59 (42%)
       New methadone start15 (27%)30 (21%)
       New buprenorphine start1 (2%)5 (4%)
       Prior MOUD continued13 (24%)46 (33%)
      DMARDs = disease-modifying antirheumatic drugs; ICU = intensive care unit; IQR = interquartile range; MOUD = medication for opioid use disorder; MRSA = methicillin-resistant Staphylococcus aureus; PICC = peripherally inserted central catheter.
      low asterisk Also known as against medical advice.
      P values were calculated using χ2 tests unless otherwise indicated.
      P value calculated using Wilcoxon rank sum test.
      § Immunosuppressed defined as those with HIV, chronic exposure to daily prednisone 10 mg or more, biologic agents or DMARDs, hematologic malignancy, or chemotherapy.
      Diagnosed with alcohol use disorder or noted to consume more than 3 or 4 alcoholic drinks daily in women and men, respectively.
      # Of people eligible for PICC placement (ie, excluding people with preexisting central access).
      low asterisklow asterisk Of those who reported opioid use.
      Infection type (bacteremia, endocarditis, vertebral osteomyelitis, and epidural abscess) and drug resistance pattern (MRSA vs MSSA) were similar between groups (Table 1). Intensive care unit admission was less common among patient-directed discharge compared with standard discharge patients (18% vs 40%). Receipt of surgical source control of infection was also less common among patient-directed discharge patients (41% vs 56%). Fewer patient-directed discharge compared with standard discharge patients received peripherally inserted central catheters (PICCs; 51% vs 71%; P < .01). Finally, in both patient-directed discharge and standard discharge patients using opioids, initiation of methadone or buprenorphine occurred in less than half of the patient-directed discharge group (47%) and standard discharge group (42%).
      Unadjusted outcomes of invasive S. aureus infections among persons who use drugs with a patient-directed discharge are shown in Table 2. Only 11% of planned antibiotic treatment courses were completed by persons who use drugs with patient-directed discharge, compared with 88% antibiotic completion in the standard discharge group (P < .01). In total, 44 of 80 (55%) of persons who use drugs with patient-directed discharge were readmitted to the hospital ≤30 days after discharge compared with 47 of 260 (18%) of persons who use drugs with standard discharges (P < .01). Overall, 39 of 44 (89%) of 30-day readmissions after patient-directed discharge were for persistent or recurrent S. aureus infection, compared with only 20 of 47 (43%) of 30-day readmissions following standard discharge. In a multivariable model adjusted for age, race/ethnicity, Charlson comorbidity score, and housing status, adjusted odds of readmission for infection in the patient-directed discharge group compared with standard discharge group was 3.8 (95% confidence interval [CI] 2.2-6.6, Table 3).
      Table 2Unadjusted Outcomes of Invasive Staphylococcus aureus Infection in Persons Who Use Drugs by Discharge Status
      Unadjusted OutcomesPatient-Directed Discharge* N = 80Standard Discharge N = 260P Value
      P values were calculated using χ2 tests unless otherwise indicated.
      Completed planned antibiotic treatment (%, n)9 (11%)230 (88%)<.01
      30-day hospital readmission (overall)44 (56%)47 (18%)<.01
      30-day hospital readmission for infection39 (49%)20 (8%)<.01
      1-year hospital readmission for infection42 (53%)57 (23%)<.01
      1-year all-cause mortality
      P value calculated using Fisher exact test. Category consists of those who survived to hospital discharge and not on hospice.
      12 (15%)28 (11%).02
      *Also known as against medical advice.
      P values were calculated using χ2 tests unless otherwise indicated.
      P value calculated using Fisher exact test. Category consists of those who survived to hospital discharge and not on hospice.
      Table 3Odds of Readmission Among Persons Who Use Drugs for Persistent or Recurrent Infection or Mortality 1 Year After Patient-Directed Discharge or Standard Discharge
      1-year hospital readmission for infection OR (95% CI)1-year all-cause mortality OR (95% CI)
      Unadjusted4.0 (2.3-6.7)1.5 (0.7-3.0)
      Adjusted for age, race/ethnicity, Charlson Comorbidity score, and housing status3.8 (2.2-6.7)1.5 (0.7-3.3)
      CI = confidence interval; OR = odds ratio.
      Cumulative mortality by 1 year following discharge was 15% among patient-directed discharge group compared with 11% in the standard discharge group (P = .02). Although unadjusted mortality was higher in the patient-directed discharge group, adjusted analysis showed odds of death were not different between groups (odds ratio 1.5 [0.7-3.3], Table 3). Of the 12 deaths in the patient-directed discharge group, none was due to S. aureus or other bacterial infection. More than half of deaths (7 of 12, 58%) were due to drug overdose. Of the 28 deaths in the standard discharge group, 5 of 28 (18%) were due S. aureus infection, and 4 of 28 (14%) were due to drug overdose.
      Review of the top causes of patient-directed discharge revealed five categories: 1) housing concerns or caring for dependents (n = 13); 2) social isolation in the hospital (n = 11); 3) being suspected of using drugs in the hospital (n = 11); 4) stigmatizing behaviors by hospital providers and staff (room searches, video surveillance, and other restrictive measures, n = 11); and 5) undertreated pain (n = 7). In 24 patients, no reason for patient-directed discharge was available in hospital records.

      Discussion

      In a large cohort of persons who use drugs hospitalized with an invasive S. aureus infection, patient-directed discharge was higher among patients with younger age, persons using injection drugs (particularly opioids and methamphetamine), and persons experiencing homelessness. Patient-directed discharge was associated with almost 4-fold increased odds of readmission within 1 year for infection compared with persons who use drugs with a standard discharge. All-cause mortality was similarly high in both discharge groups (more than 1 in 10 persons who use drugs died within a 1 year of S. aureus infection hospitalization), but most deaths in the patient-directed discharge group were due to drug overdose rather than untreated infection. Most commonly documented reasons for patient-directed discharge included logistic demands outside of the hospital and stigmatizing experiences during hospitalization.
      Thirty-day readmission rates following patient-directed discharge during infection treatment are consistently elevated across studies,
      • Kumar N.
      Burden of 30-day readmissions associated with discharge against medical advice among inpatients in the United States.
      ,
      • Alfandre D
      • Yang J
      • Harwood K
      • et al.
      “Against medical advice” discharges among HIV-infected patients: health and health services outcomes.
      • Merchant E
      • Burke D
      • Shaw L
      • et al.
      Hospitalization outcomes of people who use drugs: one size does not fit all.
      • Rapoport AB
      • Fine DR
      • Manne-Goehler JM
      • Herzig SJ
      • Rowley CF.
      High inpatient health care utilization and charges associated with injection drug use–related infections: a cohort study, 2012–2015.
      indicating the potential for ineffective treatment. Our findings of high 30-day and 1-year readmission rates among persons who use drugs with patient-directed discharge are consistent with existing data. Additionally, we found that readmission for S. aureus infection among persons who use drugs was <10% in the first month after standard discharge but almost 50% in the month following patient-directed discharge. Although we are unable to definitively distinguish between ongoing and recurrent infection, this temporality suggests that a priority issue to address is interrupted antibiotic treatment.
      Developing more effective, patient-centered treatment plans may be effective in averting patient-directed discharge and antibiotic treatment interruption. Important options include home-based outpatient parenteral antibiotics, known to be preferred by patients
      • Norris AH
      • Shrestha NK
      • Allison GM
      • et al.
      2018 Infectious Diseases Society of America Clinical Practice Guideline for the Management of Outpatient Parenteral Antimicrobial Therapy.
      yet heavily underused in persons who use drugs,
      • Ceniceros AG
      • Shridhar N
      • Fazzari M
      • Felsen U
      • Fox AD.
      Low use of outpatient parenteral antimicrobial therapy for drug use-associated infective endocarditis in an urban hospital system.
      or oral antibiotic regimens when appropriate. Marks et al
      • Marks LR
      • Liang SY
      • Muthulingam D
      • et al.
      Evaluation of partial oral antibiotic treatment for persons who inject drugs and are hospitalized with invasive infections.
      demonstrated that prescription of oral antibiotics at time of patient-directed discharge during hospitalization for infection was associated with decreased readmission rates compared to patient-directed discharge without oral antibiotics. Relatedly, expanding literature supports oral antibiotic regimens for many deep-seated infections.
      • Iversen K
      • Ihlemann N
      • Gill SU
      • et al.
      Partial oral versus intravenous antibiotic treatment of endocarditis.
      ,
      • Li H-K
      • Rombach I
      • Zambellas R
      • et al.
      Oral versus intravenous antibiotics for bone and joint infection.
      However, persons who use drugs are often excluded from these studies, which has limited their generalizability. Our study suggests that targeting antibiotic treatment interruption with innovative care delivery models might offer a pathway to reduced readmission rates and foster more patient-centered care for persons who use drugs.
      Despite the substantial morbidity associated with patient-directed discharge, we did not observe an increased 1-year all-cause mortality among patient-directed discharge patients compared with standard discharge patients. No deaths in the patient-directed discharge group were caused by untreated S. aureus infection; most were due to drug overdose. Our findings are consistent with mixed literature on this topic. Two large cohort studies of hospital discharges in the Veteran's Administration system
      • Glasgow JM
      • Vaughn-Sarrazin M
      • Kaboli PJ.
      Leaving against medical advice (ama): risk of 30-day mortality and hospital readmission.
      and in 1 urban center
      • Southern WN
      • Nahvi S
      • Arnsten JH.
      Increased risk of mortality and readmission among patients discharged against medical advice.
      demonstrated increased odds of 30-day mortality following patient-directed discharge compared with standard discharge. However, other studies specific to infection (eg, drug use-associated endocarditis) have demonstrated either similar or reduced mortality associated with patient-directed discharge.
      • Rodger L
      • Glockler-Lauf SD
      • Shojaei E
      • et al.
      Clinical characteristics and factors associated with mortality in first-episode infective endocarditis among persons who inject drugs.
      • Rudasill SE
      • Sanaiha Y
      • Mardock AL
      • et al.
      Clinical outcomes of infective endocarditis in injection drug users.
      • Eaton EF
      • Westfall AO
      • McClesky B
      • et al.
      In-hospital illicit drug use and patient-directed discharge: barriers to care for patients with injection-related infections.
      It is likely that lower age and lower frequency of comorbidities and critical illness in the patient-directed discharge group have mitigated mortality. That said, we found that 1-year all-cause mortality was strikingly high in both groups of persons who use drugs, particularly considering our population was relatively young and had relatively few comorbid illnesses. Among deaths in the patient-directed discharge group, more than half (58%) were due to fatal drug overdose, and no deaths were due to untreated S. aureus infection. Our findings indicate that, in addition to formulating effective antibiotic treatment plans with patients, addressing substance use disorders during hospitalization is of utmost importance. Hospitalization represents a “teachable moment” for patients interested in changing their substance use;
      • Velez CM
      • Nicolaidis C
      • Korthuis PT
      • Englander H.
      ‘It's been an experience, a life learning experience’: a qualitative study of hospitalized patients with substance use disorders.
      in addition, fatal overdose risk is known to be elevated following hospitalization.
      • Krawczyk N
      • Schneider KE
      • Eisenberg MD
      • et al.
      Opioid overdose death following criminal justice involvement: Linking statewide corrections and hospital databases to detect individuals at highest risk.
      Reasons for patient-directed discharge in our study raise questions about the nature and patient-centeredness of hospital care, suggesting opportunities for improvement. First, the burden of competing life priorities, such as maintaining shelter or caring for dependents and who experience overlapping challenges of mental health conditions, poverty, homelessness, and structural racism, may interfere with the ability to complete prolonged parenteral antibiotic therapy. This is another call for tailoring antibiotic delivery to the unique needs of persons who use drugs. Second, our data suggest that stigma toward persons who use drugs, including room searches or video surveillance, need to be addressed as a cause for patient-directed discharge. Our data are consistent with qualitative studies of patient perspectives on patient-directed discharge.
      • Schaefer GR
      • Matus H
      • Schumann JH
      • et al.
      Financial responsibility of hospitalized patients who left against medical advice: medical urban legend?.
      • Simon R
      • Snow R
      • Wakeman S.
      Understanding why patients with substance use disorders leave the hospital against medical advice: a qualitative study.
      • McNeil R
      • Small W
      • Wood E
      • Kerr T.
      Hospitals as a ‘risk environment’: an ethno-epidemiological study of voluntary and involuntary discharge from hospital against medical advice among people who inject drugs.
      Simon et al
      • Simon R
      • Snow R
      • Wakeman S.
      Understanding why patients with substance use disorders leave the hospital against medical advice: a qualitative study.
      cited stigma and discrimination by hospital staff as a common contributor to patient-directed discharge, along with undertreated withdrawal, substance cravings, untreated pain, and restrictive policies. Although much existing literature on patient-directed discharge identifies patient characteristics that raise the risk of patient-directed discharge, our data emphasize the need to also identify reasons for patients leaving that are rooted in processes of health care systems.
      Our study was subject to certain limitations. First, this is a retrospective study of persons who use drugs admitted to 2 San Francisco hospitals, so regional patterns of drug use and local or institutional factors may limit generalizability to other locations. Second, comparative data on care utilization (eg, receipt of peripherally inserted central catheter line, surgery, etc.) between 2 groups with different lengths of stay should be interpreted with caution. In many cases, early patient-directed discharge patients did not have the opportunity to be offered surgery. Conversely, not offering some aspects of care may have contributed to a discharge. Finally, the reported reasons for patient-directed discharge were recorded based on chart review rather than direct qualitative data directly obtained from patients. As such, information may be incomplete and may not capture the full range of patient and provider perspectives.

      Conclusion

      In conclusion, among persons who use drugs hospitalized with invasive S. aureus infection, patient-directed discharge was associated with markedly increased odds of 1-year readmission. All-cause mortality was relatively similar between patient-directed discharge and standard discharge groups, with a stark toll of drug overdose deaths in the patient-directed discharge group. Among persons who use drugs who are younger, often using opioids and methamphetamine, and experiencing homelessness, there is an urgent and growing need to develop, test, and implement novel patient-centered therapeutic strategies that prevent patient-directed discharges, allow for completion of infection therapy, and mitigate morbidity when patient-directed discharges occur.

      References

        • Ronan MV
        • Herzig SJ.
        Hospitalizations related to opioid abuse/dependence and associated serious infections from 2002 to 2012.
        Health Aff (Millwood). 2016; 35: 832-837
        • Schranz AJ
        • Fleischauer A
        • Chu VH
        • Wu L-T
        • Rosen DL.
        Trends in drug use–associated infective endocarditis and heart valve surgery, 2007 to 2017.
        Ann Intern Med. 2018; 170: 31-40
        • McCarthy NL
        • Baggs J
        • See I
        • et al.
        Bacterial infections associated with substance use disorders, large cohort of united states hospitals, 2012–2017.
        Clin Infect Dis. 2020; 71: e37-e44
        • Jackson KA
        • Bohm MK
        • Brooks JT
        • et al.
        Invasive methicillin-resistant Staphylococcus aureus infections among persons who inject drugs — six sites, 2005–2016.
        MMWR Morb Mortal Wkly Rep. 2018; 67: 625-628
        • Suzuki J
        • Robinson D
        • Mosquera M
        • et al.
        Impact of medications for opioid use disorder on discharge against medical advice among people who inject drugs hospitalized for infective endocarditis.
        Am J Addict. 2020; 29: 155-159
        • Ti L
        • Ti L.
        Leaving the hospital against medical advice among people who use illicit drugs: a systematic review.
        Am J Public Health. 2015; 105: e53-e59
        • Kim J-H
        • Fine DR
        • Li L
        • et al.
        Disparities in United States hospitalizations for serious infections in patients with and without opioid use disorder: a nationwide observational study.
        PLoS Med. 2020; 17e1003247
      1. Kimmel SD, Kim J-H, Kalesan B, Samet JH, Walley AY, Larochelle MR. Against medical advice discharges in injection and non-injection drug use-associated infective endocarditis: a nationwide cohort study [e-pub ahead of print]. Clin Infect Dis. doi: 10.1093/cid/ciaa1126. Accessed March 3, 2021.

        • Alfandre D
        • Schumann JH.
        What is wrong with discharges against medical advice (and how to fix them).
        JAMA. 2013; 310: 2393
        • Choi M
        • Kim H
        • Qian H
        • Palepu A.
        Readmission rates of patients discharged against medical advice: a matched cohort study.
        PLoS One. 2011; 6: e24459
        • Glasgow JM
        • Vaughn-Sarrazin M
        • Kaboli PJ.
        Leaving against medical advice (ama): risk of 30-day mortality and hospital readmission.
        J Gen Intern Med. 2010; 25: 926-929
        • Kumar N.
        Burden of 30-day readmissions associated with discharge against medical advice among inpatients in the United States.
        Am J Med. 2019; 132: 708-717.e4
        • Southern WN
        • Nahvi S
        • Arnsten JH.
        Increased risk of mortality and readmission among patients discharged against medical advice.
        Am J Med. 2012; 125: 594-602
        • Yong TY
        • Fok JS
        • Hakendorf P
        • Ben-Tovim D
        • Thompson CH
        • Li JY.
        Characteristics and outcomes of discharges against medical advice among hospitalised patients.
        Intern Med J. 2013; 43: 798-802
      2. Appa A, Adamo M, Le S, et al. Comparative one-year outcomes of invasive Staphylococcus aureus infections among persons with and without drug use: an observational cohort study [e-pub ahead of print]. Clin Infect Dis. doi: https://doi.org/10.1093/cid/ciab367, accessed June 2, 2021.

      3. California Department of Public Health, Center for Health Statistics. California Electronic Death Registration. Available at: https://ca.edrs.us/edrs/secureLogin. Accessed June 24, 2021.

        • Alfandre D
        • Yang J
        • Harwood K
        • et al.
        “Against medical advice” discharges among HIV-infected patients: health and health services outcomes.
        J Assoc Nurses AIDS Care. 2017; 28: 95-104
        • Merchant E
        • Burke D
        • Shaw L
        • et al.
        Hospitalization outcomes of people who use drugs: one size does not fit all.
        J Subst Abuse Treat. 2020; 112: 23-28
        • Rapoport AB
        • Fine DR
        • Manne-Goehler JM
        • Herzig SJ
        • Rowley CF.
        High inpatient health care utilization and charges associated with injection drug use–related infections: a cohort study, 2012–2015.
        Open Forum Infect Dis. 2021; 8: ofab009https://doi.org/10.1093/ofid/ofab009
        • Norris AH
        • Shrestha NK
        • Allison GM
        • et al.
        2018 Infectious Diseases Society of America Clinical Practice Guideline for the Management of Outpatient Parenteral Antimicrobial Therapy.
        Clin Infect Dis. 2019; 68: e1-e35
        • Ceniceros AG
        • Shridhar N
        • Fazzari M
        • Felsen U
        • Fox AD.
        Low use of outpatient parenteral antimicrobial therapy for drug use-associated infective endocarditis in an urban hospital system.
        Open Forum Infect Dis. 2021; 8: ofab083
        • Marks LR
        • Liang SY
        • Muthulingam D
        • et al.
        Evaluation of partial oral antibiotic treatment for persons who inject drugs and are hospitalized with invasive infections.
        Clin Infect Dis. 2020; 71: e650-e656
        • Iversen K
        • Ihlemann N
        • Gill SU
        • et al.
        Partial oral versus intravenous antibiotic treatment of endocarditis.
        N Engl J Med. 2019; 380: 415-424
        • Li H-K
        • Rombach I
        • Zambellas R
        • et al.
        Oral versus intravenous antibiotics for bone and joint infection.
        N Engl J Med. 2019; 380: 425-436
        • Rodger L
        • Glockler-Lauf SD
        • Shojaei E
        • et al.
        Clinical characteristics and factors associated with mortality in first-episode infective endocarditis among persons who inject drugs.
        JAMA Netw Open. 2018; 1e185220
        • Rudasill SE
        • Sanaiha Y
        • Mardock AL
        • et al.
        Clinical outcomes of infective endocarditis in injection drug users.
        J Am Coll Cardiol. 2019; 73: 559-570
        • Eaton EF
        • Westfall AO
        • McClesky B
        • et al.
        In-hospital illicit drug use and patient-directed discharge: barriers to care for patients with injection-related infections.
        Open Forum Infect Dis. 2020; 7: ofaa074https://doi.org/10.1093/ofid/ofaa074
        • Velez CM
        • Nicolaidis C
        • Korthuis PT
        • Englander H.
        ‘It's been an experience, a life learning experience’: a qualitative study of hospitalized patients with substance use disorders.
        J Gen Intern Med. 2017; 32: 296-303
        • Krawczyk N
        • Schneider KE
        • Eisenberg MD
        • et al.
        Opioid overdose death following criminal justice involvement: Linking statewide corrections and hospital databases to detect individuals at highest risk.
        Drug Alcohol Depend. 2020; 213107997
        • Schaefer GR
        • Matus H
        • Schumann JH
        • et al.
        Financial responsibility of hospitalized patients who left against medical advice: medical urban legend?.
        J Gen Intern Med. 2012; 27: 825-830
        • Simon R
        • Snow R
        • Wakeman S.
        Understanding why patients with substance use disorders leave the hospital against medical advice: a qualitative study.
        Subst Abus. 2019; 41: 519-525
        • McNeil R
        • Small W
        • Wood E
        • Kerr T.
        Hospitals as a ‘risk environment’: an ethno-epidemiological study of voluntary and involuntary discharge from hospital against medical advice among people who inject drugs.
        Soc Sci Med. 2014; 105: 59-66