ANALYSIS OF 2017 INMATE DEATH REVIEWS IN THE CALIFORNIA CORRECTIONAL HEALTHCARE SYSTEM Kent Imai, MD Consultant to the California Prison Receivership 11/21/2018 Analysis of 2017 CCHCS Death Reviews TABLE OF CONTENTS I. INTRODUCTION ........................................................................................................................ 1 II. DEATH REVIEW PROCESS .......................................................................................................... 1 III. DEFINITIONS ............................................................................................................................. 2 IV. Taxonomy of Care Lapses ......................................................................................................... 3 V. The California Prison Population in 2017 .................................................................................. 4 VI. Study findings ........................................................................................................................... 4 A. Number and Causes of Inmate Death with Preventability Status, 2017 ................................................................................ 4 B. Life Expectancy in the CCHCS, 2017 ....................................................................................................................................... 7 C. Not preventable Deaths in 2017............................................................................................................................................. 8 D. (Definitely) Preventable Deaths in 2017 ................................................................................................................................. 8 E. Possibly Preventable Deaths in 2017 ...................................................................................................................................... 9 F. The Taxonomy for Care Lapses in 2017 ................................................................................................................................11 G. Preventable Deaths Attributed to Lapses by Contracted Specialists and Outside Facilities ................................................12 VII. Discussion of Trends ............................................................................................................... 13 A. Trends in Prison Mortality Rates in California and the United States .................................................................................. 13 B. Trends in Specific Causes of Mortality: Top Causes..............................................................................................................14 C. Trends in Specific Causes of Mortality: End Stage Liver Disease and Liver Cancer from 2008–2016 ...................................15 D. Trends in Specific Causes of Mortality: Suicide ....................................................................................................................17 E. Trends in Specific Causes of Mortality: Homicide ................................................................................................................18 F. Trends in All Preventable Deaths, 2006–2017......................................................................................................................19 G. Trends in Care Lapses ...........................................................................................................................................................21 1. The Relationship Between the Number of Lapses and Patient Complexity ................................................ 21 2. The Relationship Between Number of Lapses and Preventability of Death in 2017 .................................. 23 VIII. Targeted Opportunities for Improvement ............................................................................... 25 A. The Primary Care Model and Preventable Deaths ...............................................................................................................25 i Analysis of 2017 CCHCS Death Reviews B. Trends in Specific Targeted Causes for Preventable Death ..................................................................................................26 IX. 1. Preventable Cardiovascular Death Rates .................................................................................................... 27 2. Preventable End Stage Liver Disease (including liver cancer) Death Rates................................................. 27 3. Preventable (Non-liver) Cancer Death Rates............................................................................................... 28 4. Deaths from Drug Overdose ........................................................................................................................ 28 5. Coccidioidomycosis Death Rates ................................................................................................................. 29 Performance Improvement Plans and Tools ........................................................................... 31 A. The CCHCS Statewide Performance Improvement Plan 2016–2018 ....................................................................................31 B. CCHCS Care Guides ...............................................................................................................................................................31 C. The Electronic Medical Record .............................................................................................................................................32 X. Conclusions............................................................................................................................. 32 ii Analysis of 2017 CCHCS Death Reviews LIST OF TABLES AND FIGURES TABLE 1. CAUSES OF DEATH AND PREVENTABILITY STATUS AMONG ALL CALIFORNIA INMATES, 2017. ................................... 5 TABLE 2. TOP CAUSES OF DEATH AMONG CALIFORNIA INMATES, 2017, COMPARED TO AMERICAN MALE DEATHS, 2015 (MOST RECENT DATA AVAILABLE). ....................................................................................................................................... 7 TABLE 3. RANGES AND AVERAGE AGES AT DEATH AMONG ALL CALIFORNIA INMATES, 2017 ................................................... 8 FIGURE 1. TREND IN CCHCS DEFINITELY PREVENTABLE DEATHS, 2006–2017............................................................................. 8 TABLE 4. CAUSES OF POSSIBLY PREVENTABLE DEATH AMONG CALIFORNIA INMATES, 2017. ................................................... 9 TABLE 5. SUMMARY OF CARE LAPSES, 2017.............................................................................................................................. 11 FIGURE 2. POSSIBLY PREVENTABLE DEATHS OF CALIFORNIA PRISON INMATES INVOLVING LAPSES BY CONTRACTED SPECIALISTS OR OUTSIDE FACILITIES, 2008–2017. ............................................................................................................. 12 TABLE 6. ANNUAL DEATH RATES AMONG CALIFORNIA AND U.S. STATE PRISON INMATES, 2006–2017. ................................ 13 FIGURE 3. TRENDED DEATH RATE PER 100,000 INMATES, CCHCS AND TOTAL U.S. STATE PRISONS, 2006–2017. ................... 14 TABLE 7. TOP CAUSES OF DEATH AMONG CALIFORNIA INMATES, 2006–2017. ....................................................................... 14 TABLE 8. CCHCS CHRONIC HEPATITIS C ASSOCIATED DEATHS, 2008–2017. ............................................................................. 16 FIGURE 4. CCHCS CHRONIC HEPATITIS C ASSOCIATED DEATH RATES, 2008–2017. .................................................................. 16 TABLE 9. NUMBERS AND RATES OF SUICIDE-RELATED DEATHS: CALIFORNIA, ALL U.S. STATE PRISONS, 2006–2017. ............. 17 FIGURE 5. SUICIDE DEATH RATES IN THE CALIFORNIA CORRECTIONAL SYSTEM (2006-2017) AND U.S. STATE PRISONS (2006–2014). ...................................................................................................................................................................... 17 TABLE 10. NUMBERS OF HOMICIDE-RELATED DEATHS IN CALIFORNIA AND ALL U.S. STATE PRISONS, 2006–2017. ............... 18 FIGURE 6. HOMICIDE DEATH RATES IN THE CALIFORNIA CORRECTIONAL SYSTEM AND ALL U.S. STATE PRISONS, 2006– 2017. ................................................................................................................................................................................... 19 TABLE 11. RATES OF PREVENTABLE DEATHS AMONG CALIFORNIA INMATES, 2006–2017. ...................................................... 19 FIGURE 7. NUMBER OF PREVENTABLE DEATHS IN THE CALIFORNIA CORRECTIONAL HEALTHCARE SYSTEM, 2006-2017. ..... 20 FIGURE 8. PREVENTABLE DEATH RATES IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017......................................... 20 FIGURE 8A. THREE PERIODS IN PREVENTABILITY TRENDS, CCHCS 2006-2017. ......................................................................... 21 TABLE 12. FREQUENCY OF ASSOCIATED CONDITIONS (EXCLUSIVE OF PRIMARY CAUSE OF DEATH) IN CCHCS INMATE DEATHS, 2015 AND 2016. ................................................................................................................................................... 21 TABLE 13. NUMBER OF LAPSES BY CATEGORY OF PREVENTABILITY, 2017. .............................................................................. 23 FIGURE 9. AVERAGE NUMBER OF LAPSES PER CASE BY PREVENTABILITY, 2007–2017. ............................................................ 23 iii Analysis of 2017 CCHCS Death Reviews TABLE 14. NUMBER OF LAPSES, BY PREVENTABILITY, IN CCHCS DEATHS, 2007–2017. ............................................................ 24 FIGURE 10. TREND IN ANNUAL AVERAGE OF CARE LAPSES PER DEATH, CCHCS, 2007–2017. .................................................. 24 TABLE 15. IDENTIFIABLE PRIMARY CARE IN CALIFORNIA INMATE DEATH CASES, 2009–2017. ................................................ 25 FIGURE 11. PERCENTAGE OF DEATHS IN THE CCHCS WITH AN IDENTIFIED PRIMARY CARE PHYSICIAN, AND CORRESPONDING RATES OF PREVENTABLE DEATH, 2009–2017. ...................................................................................... 26 TABLE 16. NUMBERS AND RATES OF PREVENTABLE DEATHS FROM CARDIOVASCULAR, END STAGE LIVER DISEASE, AND CANCER IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017. .................................................................................. 26 FIGURE 12. PREVENTABLE CARDIOVASCULAR DEATHS – NUMBER OF PREVENTABLE CASES AND RATES OF DEATH IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017. .................................................................................................... 27 FIGURE 13. PREVENTABLE END STAGE LIVER DISEASE DEATHS – NUMBER OF PREVENTABLE CASES AND RATES OF DEATHS IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017. .................................................................................. 27 FIGURE 14. PREVENTABLE CANCER DEATHS – NUMBER OF PREVENTABLE CASES AND RATES OF DEATH IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017............................................................................................................. 28 TABLE 17. NUMBERS AND RATES OF DRUG OVERDOSE-RELATED DEATHS IN THE CALIFORNIA CORRECTIONAL HEALTHCARE SYSTEM AND IN ALL U.S. PRISONS, 2006–2017............................................................................................ 28 FIGURE 15. DRUG OVERDOSE DEATH RATES IN THE CALIFORNIA CORRECTIONAL HEALTHCARE SYSTEM, 2006–2017. ......... 29 TABLE 18. COCCIDIOIDOMYCOSIS RELATED DEATHS IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017. ................... 30 FIGURE 16. COCCIDIOIDOMYCOSIS RELATED DEATHS AND DEATH RATES IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017. ......................................................................................................................................................................... 30 iv Analysis of 2017 CCHCS Death Reviews I. INTRODUCTION The California Correctional Healthcare System (CCHCS) was placed under Federal Receivership in October 2005, when the State of California was found to be violating State prisoner rights under the eighth amendment to the U.S. Constitution. Medical care in the state’s prisons was so poor that an estimated average of one inmate each week died as a result of malpractice or neglect. Since that time, the Receivership has been transforming the CCHCS in order to provide constitutionally adequate medical care to the inmates in the 35 prison facilities. The CCHCS website, cchcs.ca.gov, highlights the mission of the Receiver: • to reduce unnecessary morbidity and mortality and protect public health by providing timely access to safe, efficient, coordinated medical care; and • to move from a system of chaotic care that was largely episodic to a system of proactive, planned, informed, patient-centered and professional care. This is the twelfth annual analysis of inmate death reviews in the CCHCS. It describes the CCHCS death review process and how it is intended to reduce the occurrence of preventable deaths. All causes of death, serious care lapses and preventable deaths are identified and trended from 2006 through 2017. This and all prior death report analyses are available at https://cchcs.ca.gov/reports/. II. DEATH REVIEW PROCESS The CCHCS maintains a Death Reporting and Review Program in which every patient death occurring within the custody of the California Department of Corrections and Rehabilitation (CDCR) is reviewed. The purpose of the program is to identify patterns of lapses in care related to the cause of death and to determine opportunities for improvement in the delivery of health care. When an inmate death occurs, an initial death review summary is submitted within five calendar days to the statewide Death Review Unit (DRU) by the institution where the death occurred. This initial report includes a chronology of significant events including the emergency medical response, any identified lapses in health care delivery and any identified system issue which may have contributed to death. At the DRU, each death is assigned to a physician reviewer and a nurse reviewer. An extensive review of the patient’s medical and nursing care is conducted. Every clinical encounter in the six months prior to death, and if relevant, beyond six months, is reviewed. The quality of care experienced by the patient at each encounter is evaluated. Factors evaluated include the quality of triage and evaluation, timeliness of access to care, the quality of care for any chronic medical condition, adherence to published evidence-based care guides, responses to all abnormal laboratory and X-ray studies, and the timing and quality of emergency response. In addition, the presence of a primary care physician and adherence to a primary care model of care delivery is noted. 1 Analysis of 2017 CCHCS Death Reviews All suicides or possible suicides undergo a separate case review by a member of the Suicide Prevention and Response Focused Improvement Team (SPRFIT), which includes a Mental Health Program review. In every case, the cause of death is determined. Care lapses are noted, especially any that may have contributed to the patient death. The physician reviewer then makes a judgment as to whether the death was preventable, possibly preventable or not preventable. Each death review is presented by the assigned reviewer to the Death Review Committee (DRC). The DRC membership is appointed by the Statewide Deputy Directors of Medical and Nursing Services. The DRC consists of three physicians, three nurses, one mental health professional, one custody representative, and one (non-voting) member of the Quality Management staff. The DRC is co-chaired by a physician and nurse executive member. Following discussion of the case, the DRC votes to attribute cause of death and the level of preventability. Functions of the death review process include identifying individual providers for further peer review, identifying opportunities for improvement in healthcare policies and practices, making recommendations for changes to existing interdisciplinary care guides, and highlighting systemic areas in need of improvement. Extreme departures from the standard of care are referred to the Medical, Nursing or Mental Health Peer Review Committees or in the case of any sentinel event, to the Patient Safety Program. A major purpose of the death review process is to reduce the occurrence of preventable death. III. DEFINITIONS The following definitions are taken from the Inmate Medical Services Policies and Procedures, Volume 1, Chapter 29.1, Death Reporting and Review Program Policy, and are used in this report. Expected Death: A medically anticipated death which is related to the natural course of a patient’s illness or underlying condition. Unexpected Death: Any unanticipated death which is not related to the natural course of a patient’s illness or underlying condition. Extreme Departure: Care given that may cause injury or expose patients to some substantial risk of injury or harm which no other reasonable and competent provider would provide under the same or similar circumstances. Not Preventable Death: A death that could not have been prevented or significantly delayed despite identified opportunities for improvement in the medical care or systemic issues. Possibly Preventable Death: A death wherein opportunities for clinical intervention or significant lapses related to care delivery have been identified that may have prevented or significantly delayed the patient’s death. Preventable Death: A death wherein opportunities for clinical intervention or significant lapses related to care delivery have been identified that would have prevented or significantly delayed the patient’s death. Care lapse: Any departure from the standard of care which poses a risk to patient safety. 2 Analysis of 2017 CCHCS Death Reviews IV. TAXONOMY OF CARE LAPSES In 2008, a taxonomy of types of medical errors or care lapses was incorporated into this annual review and was used to organize the findings of the DRC reviewers. When used systematically, this taxonomy has proven to be a useful quality improvement tool for identifying the common reasons for substandard healthcare that might result in preventable deaths. It has been useful for identifying potentially unsafe clinical practice, opportunities for system and process redesign, and educational strategies for CCHCS clinical staff. In the 2008 taxonomy, care lapses are organized into fourteen separate types. Type 1 – Failure to recognize, evaluate and manage important symptoms and signs – so called clinical “red flags.” Type 2 – Failure to follow clinical care guides or departmental policies developed and endorsed by the medical department of the CCHCS. These include evidence-based guidelines for the management of asthma, diabetes mellitus, hepatitis C infection, HIV/AIDS, chronic pain, and care at the end of life. Other care guides outline standards for the management of hypertension, acute coronary syndromes, congestive heart failure, cardiac arrhythmia, and anticoagulation. Type 3 – Delay in access to the appropriate level of care, of sufficient duration as to result in harm to the patient. Type 4 – Failure to identify and appropriately respond to abnormal test results. Type 5 – Failure of appropriate communication between providers, especially at points where transfers of care occur (care transitions). Type 6 – Fragmentation of care resulting from failure of an individual clinician or the primary care team to assume responsibility for the patient’s care - lack of a primary care model. Type 7 – Iatrogenic injury resulting from a surgical or procedural complication. Type 8 – Medication prescribing error, including failure to prescribe an indicated medication, failure to do appropriate monitoring, or failure to recognize and avoid known drug interactions. Type 9 – Medication delivery error, including significant delay in a patient receiving medication or a medication delivered to the wrong patient. Type 10 – Practicing outside the scope of one’s professional capability (may apply to nursing staff, midlevel practitioners, or physicians). Type 11 – Failure to adequately supervise a midlevel practitioner, including failure to be readily available for consultation or an administrative failure to provide for appropriate supervision. Type 12 – Failure to communicate effectively with the patient. Type 13 – Patient non-adherence with suggestions for optimal care. Type 14 – Delay or failure in emergency response, including delay in activation or failure to follow the emergency response protocol. 3 Analysis of 2017 CCHCS Death Reviews In 2016, the DRC developed a new taxonomy for classifying care lapses. This new taxonomy is more detailed in capturing the various causes for errors in coordination and continuity of care, clinical management, medication management, emergency medical care, transportation, nursing encounters, and utilization management. This new taxonomy is not yet in use with the analysis of 2016 and 2017 death reviews. Both to support longitudinal analyses, and because the new taxonomy has not yet been consistently adopted by the physician and nurse reviewers in the DRC, the current review continues to use the original 2008 taxonomy. V. THE CALIFORNIA PRISON POPULATION IN 2017 At the beginning of the Receivership in 2006, prison overcrowding was identified as a major factor contributing to the delivery of substandard medical care. Between 2008 and 2015, the California Department of Corrections and Rehabilitation (CDCR) significantly reduced the prison population by 25 percent. In 2006, the number of inmates in the CCHCS was 171,310. By 2015, the average number of inmates in the CCHCS was 128,477. In 2017, the average number of total inmates was 130,807. Of those, 124,888 (95.5%) were males and 5,919 (4.5%) were females. VI. STUDY FINDINGS A. Number and Causes of Inmate Death with Preventability Status, 2017 There were 388 inmate deaths in 2017; 378 in males (97.4%) and 10 in females (2.6%). Of these, the death review committee designated 373 deaths as not preventable, 14 deaths as possibly preventable and 1 death as (definitely) preventable. Table 1 shows the causes of death and preventability status in 2017. Cancer (98 cases) was the top cause of death, with lung cancer (14 cases), colon cancer (8 cases; 7 not preventable and 1 possibly preventable) and multiple myeloma (6 cases) the top 3 types of cancer. Cardiovascular disease (68 cases) was the second leading cause of death, with sudden cardiac arrest (27 cases), congestive heart failure (17 cases; 16 not preventable and 1 possibly preventable) and acute myocardial infarction (13 cases; 11 not preventable and 2 possibly preventable) accounting for 57 cases or 84% of all cardiovascular deaths. Liver disease (41 cases) was the third most common cause of death. As in past years, we have grouped end stage liver disease and liver cancer together because liver cancer is a consequence of cirrhosis (end stage liver disease) and the two conditions almost always coexist in the same patient. As in past years, chronic hepatitis C infection was the underlying cause of end stage liver disease with or without liver cancer in the California prison population. Hepatitis C virus infected 14% of all CCHCS prisoners in 2016, and disproportionally infected patients who died in 2017 (32%). Drug overdose (40 cases) was the fourth most common cause of death in 2017. Suicide (31 cases) and homicide (19 cases) were the sixth and seventh most common reasons for death in the CCHCS in 2017. 4 Analysis of 2017 CCHCS Death Reviews Infectious disease (33 cases) was the fifth most common cause of death in 2017; this category covers a variety of different causes, including 15 cases of pneumonia, 8 cases of sepsis, 4 cases of infectious endocarditis, and single cases of cellulitis HIV/AIDS, influenza, meningitis and necrotizing fasciitis. TABLE 1. CAUSES OF DEATH AND PREVENTABILITY STATUS AMONG ALL CALIFORNIA INMATES, 2017. NUMBER OF CASES CAUSES OF DEATH NOT PREVENTABLE 98 Cancer 95 NOT PREVENTABLE: 13 Lung; 7 Colon; 6 Multiple Myeloma; 5 Bladder; 5 Kidney; 5 Prostate; 5 Unknown primary; 4 Melanoma; 4 Pancreas; 3 Bile duct; 3 Esophagus; 3 Larynx; 3 Leukemia-Acute Myelogenous; 3 Squamous Cell; 3 Stomach; 2 Lymphoma, non Hodgkin; 2 Myelodysplasia; 2 Myelofibrosis; 2 Nasopharynx; 2 Renal; 1 Anus; 1 Brain; 1 Breast; 1 Duodenum; 1 Hodgkin Lymphoma; 1 Hypopharynx; 1 Leukemia-Acute Lymphoblastic; 1 Leukemia-Acute Monocytic; 1 Mesothelioma; 1 Rectum; 1 Synovial Sarcoma; 1 Testes; 1 Tonsil 68 Cardiovascular Disease 63 NOT PREVENTABLE: 27 Sudden Cardiac Arrest; 16 Congestive Heart Failure; 11 Acute Myocardial Infarction; 2 Cardiomyopathy; 2 Coronary Artery Disease; 2 Aortic Dissection; 1 Acute myocarditis; 1 Aortic aneurysm Dissection; 1 Post-CABG complication 41 Liver Disease 41 NOT PREVENTABLE: 22 End Stage Liver Disease (incl. 1 non-HepC) ; 18 Liver Cancer; 1 Acute Hepatic Failure 40 Drug Overdose 40 NOT PREVENTABLE: 39 Non-prescribed: 11 Heroin; 11 Methamphetamine; 6 Fentanyl; 5 Unknown Opiate; 3 Opiate plus Methamphetamine; 3 Opiate plus Fentanyl. 1 Prescribed: 1 Antipsychotic 32 Infectious Disease 31 NOT PREVENTABLE: 15 Pneumonia (includes 6 aspiration pneumonia); 4 Infectious Endocarditis; 7 Septicemia; 1 Cellulitis; 1 HIV/AIDS; 1 Influenza; 1 Meningitis 31 Suicide 29 NOT PREVENTABLE PREVENTABLE/POSSIBLY PREVENTABLE 3 POSSIBLY PREVENTABLE: 1 Colon, 1 Stomach, 1 Testes 5 POSSIBLY PREVENTABLE: 2 Acute Myocardial Infarction; 2 Cardiac Arrhythmia; 1 Congestive Heart Failure 2 POSSIBLY PREVENTABLE: 1 Necrotizing cervical fasciitis; 1 Septicemia 1 PREVENTABLE 1 POSSIBLY PREVENTABLE 5 Analysis of 2017 CCHCS Death Reviews 19 Homicide 19 NOT PREVENTABLE: 19 Homicide by Inmate(s) 11 Cerebrovascular Disease 11 NOT PREVENTABLE: 6 Stroke, Ischemic; 5 Stroke, Hemorrhagic 10 Pulmonary 10 NOT PREVENTABLE: 7 Chronic Obstructive Pulmonary Disease; 3 Pulmonary Fibrosis 7 each Circulatory System 7 NOT PREVENTABLE: 5 Pulmonary Embolism; 1 Shock, Hypovolemic; 1 Third-degree Heart Block Renal Disease 7 NOT PREVENTABLE: 6 End Stage Renal Disease; 1 Chronic Renal Failure Gastrointestinal Disease 6 NOT PREVENTABLE: 3 Upper GI Hemorrhage; 1 Ischemic Bowel Syndrome; 1 Pancreatitis; 1 Ulcerative Colitis 6 Neurological Disease 5 NOT PREVENTABLE: 3 Dementia; 1 Alzheimer Dementia; 1 Colloid cyst of third ventricle 4 Trauma 3 NOT PREVENTABLE: 1 Accidental Trauma; 1 Shock, Hypovolemic; 1 Traumatic Rupture of Spleen 3 Autoimmune 3 NOT PREVENTABLE: 1 Systemic Lupus Erythematosus; 1 Mixed Connective Tissue Disease; 1 Rheumatoid Arthritis 2 Accidental Injury to Self 2 NOT PREVENTABLE 1 each Metabolic 1 NOT PREVENTABLE: General Deterioration Unknown 1 NOT PREVENTABLE: Possible Homicide 388 Total 373 NOT PREVENTABLE 1 POSSIBLY PREVENTABLE: 1 Aspiration 1 POSSIBLY PREVENTABLE: 1 Neuroleptic Malignant Syndrome 1 POSSIBLY PREVENTABLE: 1 Subdural Hematoma 1 PREVENTABLE 14 POSSIBLY PREVENTABLE 6 Analysis of 2017 CCHCS Death Reviews Table 2 compares the top causes of death in CCHCS men with those in the free living American male population. Significant differences can be seen. In the prison population, cancer (25.3%) was the number one cause of death while cardiovascular disease (17.5%) was second most frequent and liver disease (10.6%) was third. These three accounted for 53% of all deaths. For the American male population in 2015 (the last year for which statistics are available), cardiovascular disease (24.4%) was number one, cancer (22.8%) ranked number two, and accidental injury was a distant third (6.8%). Chronic liver disease accounted for 1.9% and ranked tenth. Drug overdose, infectious diseases, suicide and homicide were all significantly higher in the prison population than in free living American males. TABLE 2. TOP CAUSES OF DEATH AMONG CALIFORNIA INMATES, 2017, COMPARED TO AMERICAN MALE DEATHS, 2015 (MOST RECENT DATA AVAILABLE). CCHCS 2017 AMERICAN MALES 2015 1. Cancer (25.3%) 1. Cardiovascular (24.4%) 2. Cardiovascular (17.5%) 2. Cancer (22.8%) 3. Liver disease (end stage), includes liver cancer (10.6%) 3. Accidental injury (6.8%) 4. Drug overdose (10.3%) 4. Chronic respiratory (5.3%) 5. Infectious diseases (8.5%) 5. Stroke (4.2%) 6. Suicide (8.0%) 6. Diabetes mellitus (3.1%) 7. Homicide (4.9%) 7. Suicide (2.5%) 8. Cerebrovascular Disease (2.8%) 8. Alzheimer’s Disease (2.5%) 9. Pulmonary (2.6%) 9. Influenza and pneumonia (2.0%) 10. Renal disease (1.8%) 10. Chronic liver disease (1.9%) B. Life Expectancy in the CCHCS, 2017 The average age at death of all CCHCS male patients in 2017 was 56 years. That of females was 54 years. Non incarcerated American males and females enjoy a life expectancy some two decades longer. In 2016, the American male life expectancy was 76.3 years. Life in prison is hard and relatively short, and life expectancy appears to be bimodal. Drug overdoses, suicides and homicides cause death at an average of 38 years, whereas prisoners dying from all other causes live to an average age of 62 years. 7 Analysis of 2017 CCHCS Death Reviews TABLE 3. RANGES AND AVERAGE AGES AT DEATH AMONG ALL CALIFORNIA INMATES, 2017 AGE RANGE AVERAGE AGE Age of all 378 male decedents 21 – 91 56 Age of all 10 female decedents 32 – 82 54 Age of suicides, drug overdoses, and homicides 21 – 67 38 21 – 61 34 22 – 67 41 22 – 61 39 22 – 91 62 Drug overdose Age excluding suicide, drug overdose, and homicide C. Not preventable Deaths in 2017 The 373 deaths classified by the DRC as not preventable in 2017 were 96.1% of the total. As seen in Table 1, there were 40 drug overdoses, 29 suicides, 19 homicides, 3 cases of accidental trauma and 2 cases of accidental selfinjury. The remaining 293 deaths were the result of chronic underlying disease. D. (Definitely) Preventable Deaths in 2017 Figure 1 shows the run chart for all definitely preventable deaths in the CCHCS from 2006 through 2017. FIGURE 1. TREND IN CCHCS DEFINITELY PREVENTABLE DEATHS, 2006–2017. Definitely preventable deaths per year, CCHCS, 2006-2017. 23 18 18 14 9 3 5 0 2006 2007 5 2008 3 2009 5 2010 2 1 0 0 0 0 1 2011 2012 2013 2014 2015 2016 2017 There was one (definitely) preventable death in 2017. This is the first case since 2012 that was thought to be definitely preventable by the members of the Death Review Committee. This case is described below. A type 2 lapse – failure to follow clinical guidelines for care, and a type 14 lapse – delay or failure in emergency response – contributed to this preventable death. 8 Analysis of 2017 CCHCS Death Reviews A 28 year old man died of suicide by asphyxiation. Failure to adequately monitor the patient during suicide watch, and a subsequent twenty five minute delay in activation of 911 contributed to this preventable death. E. Possibly Preventable Deaths in 2017 There were 14 deaths classified by the DRC as possibly preventable in 2017. Table 4 shows the causes of death in these cases. TABLE 4. CAUSES OF POSSIBLY PREVENTABLE DEATH AMONG CALIFORNIA INMATES, 2017. NUMBER OF CASES CAUSE OF DEATH 5 CARDIOVASCULAR DISEASE: 2 Acute Myocardial Infarction; 2 Cardiac Arrhythmia; 1 Congestive Heart Failure 3 CANCER: 1 Colon, 1 Stomach, 1 Testes 2 INFECTIOUS: 1 Necrotizing cervical fasciitis; 1 Septicemia 1 each GASTROINTESTINAL DISEASE: Aspiration NEUROLOGICAL DISEASE: Neuroleptic Malignant Syndrome SUICIDE TRAUMA: Subdural Hematoma 14 Total Each case is described briefly below and the type of lapse most contributory to each death is noted. A type 1 lapse – failure to recognize or evaluate important symptoms or signs – was identified as a significant factor in eleven cases of possibly preventable death. 1. A 56 year old man died of sepsis secondary to a necrotizing neck infection. Failure to adequately evaluate persistent fever and dysphagia in this known diabetic led to a delay in diagnosis which contributed to his possibly preventable death. 2. A 66 year old man with post polio syndrome and a clinical picture of intestinal obstruction died of aspiration of gastric contents. Failure to properly evaluate the patient’s lethargy and distended abdomen coupled with failure to review an abnormal abdominal X-ray may have contributed to this death. 3. A 52 year old man with end stage renal failure on hemodialysis died of sudden cardiac arrest. The patient complained of an inability to move. This was not evaluated for 90 minutes, when the patient suddenly became unresponsive. Resuscitation was unsuccessful. 9 Analysis of 2017 CCHCS Death Reviews 4. A 36 year old man died of a cardiac arrhythmia in an out of state facility. The consulting cardiologist failed to aggressively evaluate new onset atrial fibrillation coupled with an abnormal echocardiogram showing severe cardiac dysfunction, and the patient died one day later in the prison to which he had been returned. 5. A 72 year old man with diabetes mellitus died of sudden cardiac arrest. Poorly managed diabetes mellitus and prolonged hypoglycemia may have contributed to his death. 6. A 62 year old man died of septic shock from acute and chronic cholecystitis. A one month delay in diagnosis was thought to result from a failure to aggressively evaluate a recurrent fever. 7. A 70 year old man died of cancer of the stomach. An incomplete evaluation of chronic weight loss caused an 8 month delay in diagnosis which possibly contributed to the patient’s death. 8. A 46 year old man died of apparent neuroleptic malignant syndrome. Incomplete evaluation of high fever and other abnormal vital signs in the setting of multiple psychoactive medications led to several days delay in diagnosis, contributing to this possibly preventable death. 9. A 73 year old man with multiple cardiac risk factors died of probable myocardial infarction. Failure to aggressively evaluate nausea, hypotension and tachycardia contributed to this cardiac death. 10. A 26 year old man died of metastatic testicular cancer. The diagnosis was delayed by 13 months because of failure to follow up an abnormal testicular examination. The resulting delay in treatment contributed to this possibly preventable death. 11. A 74 year old man died of an acute subdural hematoma and brain contusion. A one hour delay in evaluation of the patient after a spontaneous fall in the prison yard was thought to contribute to a delay in treatment which might have prevented his death. A type 2 lapse – failure to follow clinical guidelines for care – contributed to the following possibly preventable death. 12. A 60 year old man with hypertension died of congestive heart failure. Failure to regularly monitor and treat his high blood pressure during the six year period that preceded his death might have contributed to the development of congestive heart failure. A type 4 lapse – failure to identify or appropriately respond to abnormal test results – contributed to the following possibly preventable death. 13. A 71 year old man died of metastatic colon cancer. A failure to adequately evaluate anemia and a positive stool occult blood test led to an 18 month delay in diagnosis. A type 14 lapse – delay or failure in emergency response – contributed to the following possibly preventable death. 10 Analysis of 2017 CCHCS Death Reviews 14. A 31 year old man died of multiple self inflicted lacerations with resultant exsanguination. A 36 minute delay in activating 911 may have contributed to his death. F. The Taxonomy for Care Lapses in 2017 One of the primary purposes of the death reviews is to identify lapses in care, regardless of whether these lapses lead to a patient death. Recognition of lapses presents opportunities for system improvement and for targeted provider and staff education. The taxonomy for tracking these lapses has been described. Table 5 summarizes these lapses in all of the 2017 deaths. TABLE 5. SUMMARY OF CARE LAPSES, 2017. LAPSES OF CARE TYPES # OF LAPSES IN 373 NOT PREVENTABLE DEATHS # OF LAPSES IN # OF LAPSES IN 1 TOTAL 14 POSSIBLY PREVENTABLE LAPSES IN PREVENTABLE DEATH ALL 388 DEATHS DEATHS #1 – Failure to recognize, identify or adequately evaluate important symptoms or signs 29 10 #2 – Failure to follow established guidelines for evaluation and/or management of a specific condition 10 1 #3 – Delay in access to care sufficient to result in harm to the patient 6 1 7 #4 – Failure to adequately pursue abnormal test results 1 3 4 2 1 3 #5 – Failure of provider-to-provider communications including botched handoffs #6 – Fragmentation of care such that individual responsibility for patient is waived 39 1 12 3 3 #8 – Medication prescribing error 5 5 #9 – Medication delivery error 3 3 #10 – Practicing outside the scope of one’s professional capabilities 1 1 1 1 #7 – Surgical/procedural complication resulting in iatrogenic injury #11 – Unsupervised mid-level (nurse practitioner or physician assistant) care #12 – Failure to communicate effectively with the patient 11 Analysis of 2017 CCHCS Death Reviews LAPSES OF CARE TYPES # OF LAPSES IN 373 NOT PREVENTABLE DEATHS # OF LAPSES IN # OF LAPSES IN 1 TOTAL 14 POSSIBLY PREVENTABLE LAPSES IN PREVENTABLE DEATH ALL 388 DEATHS DEATHS 6 6 #13 – Patient non-adherence with recommendation for optimal care #14 – Delay in emergency response or failure to follow emergency response protocol 25 #15 – Other 16 All Types 1 1 27 16 108 17 2 127 The DRC reviewers identified 108 lapses in the 373 not preventable deaths, 17 lapses in the 14 possibly preventable deaths, and 2 lapses in the one definitely preventable death. In total, there were 127 lapses noted in the 388 death reviews. G. Preventable Deaths Attributed to Lapses by Contracted Specialists and Outside Facilities In 2017, there was one case in which a consulting specialist in an out-of-state contracted facility failed to adequately manage a patient with signs and symptoms of severe coronary artery disease, resulting in a possibly preventable death (case #4 of possibly preventable deaths, above). FIGURE 2. POSSIBLY PREVENTABLE DEATHS OF CALIFORNIA PRISON INMATES INVOLVING LAPSES BY CONTRACTED SPECIALISTS OR OUTSIDE FACILITIES, 2008–2017. Number of preventable cases involving lapses by contracted specialists or outside facilities % of preventable cases 20 30% 15 23% 10 15% 5 8% 0 0% 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 12 Analysis of 2017 CCHCS Death Reviews VII. DISCUSSION OF TRENDS A. Trends in Prison Mortality Rates in California and the United States Annual death rates in the CCHCS are shown in Table 6 and compared to rates in all U.S. state prisons. (U. S Bureau of Justice statistics, bjs.gov). The 2017 death rate is 297/100,000 inmates for CCHCS. This represents the highest annual mortality rate in the past 12 years. The average death rate from 2006–2014 for all U.S. State prisons is 260/100,000. Comparable statistics for the years 2015–2017 are not available. TABLE 6. ANNUAL DEATH RATES AMONG CALIFORNIA AND U.S. STATE PRISON INMATES, 2006–2017. YEAR CCHCS NUMBER OF DEATHS CCHCS NUMBER OF INMATES CCHCS DEATH RATE PER 100,000 INMATES TOTAL U.S. STATE PRISON DEATH RATE PER 100,000 2006 424 171,310 248 249 2007 395 170,786 231 256 2008 369 170,022 217 260 2009 393 169,459 232 257 2010 415 166,700 249 245 2011 388 161,843 240 260 2012 362 134,929 268 265 2013 366 133,297 275 274 2014 319 135,225 236 275 2015 355 128,824 276 not available 2016 334 128,705 260 not available 2017 388 130,807 297 not available 252 (217–297) 260 (245–275) Average (Range) Figure 3 below shows the trended death rates for the CCHCS from 2006–2017, and the trended death rates for all US prisons from 2006– 2014. There has been a slight increase in the rates of death both in California and in all US prisons. 13 Analysis of 2017 CCHCS Death Reviews FIGURE 3. TRENDED DEATH RATE PER 100,000 INMATES, CCHCS AND TOTAL U.S. STATE PRISONS, 2006–2017. CCHCS TOTAL U.S. State Prison 350 300 250 200 150 100 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 B. Trends in Specific Causes of Mortality: Top Causes Table 7 shows the top nine causes of death in the CCHCS from 2006–2016. Beginning in 2015, cardiovascular causes of death overtook end stage liver disease as the number 2 overall cause of death. In 2017, drug overdose deaths rose to become the fourth most common cause of death, and suicides and homicides remained as the sixth and seventh most common causes of death in the incarcerated population. Comparisons with the 2016 death review analysis show that there were 16 more cancer deaths, 16 more cardiovascular deaths, and 11 more drug overdose deaths in 2017 than in 2016. These 43 additional deaths in 2017 account for most of the difference in the mortality rates between 2017 and 2016. TABLE 7. TOP CAUSES OF DEATH AMONG CALIFORNIA INMATES, 2006–2017. YEAR RANK 1 2 3 4 5 6 7 8 9 2017 Cancer Cardiovascular Disease End Stage Liver Disease* Drug Overdose Infectious Disease** Suicide Homicide Cerebrovascular Disease Pulmonary 2016 Cancer Cardiovascular Disease End Stage Liver Disease* Infectious Disease** Drug Overdose (tied) Suicide, Homicide Cerebrovascular Disease Pulmonary 2015 Cancer Cardiovascular Disease End Stage Liver Disease* Infectious Disease** Suicide Drug Overdose Cerebrovascular Disease Pulmonary 2014 Cancer End Stage Liver Disease* Cardiovascular Disease Suicide Drug Overdose Pneumonia Homicide Pulmonary (tied) Infectious; StrokeHemorrhagic 2013 Cancer End Stage Liver Disease* Cardiovascular Disease Suicide Drug Overdose Homicide Sepsis (tied) Pulmonary; Pneumonia 2012 Cancer End Stage Liver Disease* Cardiovascular Disease Suicide Homicide Drug Overdose (tied) Sepsis; Infectious Homicide Stroke 14 Analysis of 2017 CCHCS Death Reviews YEAR RANK 2011 1 2 3 4 5 6 7 8 9 Cancer End Stage Liver Disease* Cardiovascular Disease Suicide Pneumonia Homicide Sepsis Drug Overdose Stroke Suicide (tied) Drug Overdose; Homicide Pneumonia Congestive Heart Failure (tied) Coccidioidomycosis; End Stage Renal Disease; Stroke 2010 Cancer End Stage Liver Disease* Cardiovascular Disease 2009 Cancer End Stage Liver Disease* Cardiovascular Disease Suicide Drug Overdose Pneumonia Congestive Heart Failure Homicide 2008 Cancer Suicide End Stage Liver Disease* Cardiovascular Disease Drug Overdose Pneumonia HIV/AIDS Congestive Heart Failure Sepsis 2007 Cancer* End Stage Liver Disease Cardiovascular Disease Suicide Homicide HIV/AIDS Stroke Drug Overdose Pneumonia 2006 Cancer* Cardiovascular Disease End Stage Liver Disease Suicide Drug Overdose Homicide Pulmonary End Stage Renal Disease Stroke * Liver Cancer was counted as Cancer in 2006 and 2007; as Liver Disease from 2008 onward. ** Beginning with 2015, Pneumonia and Sepsis were included in Infectious Disease, which also includes HIV/AIDS. C. Trends in Specific Causes of Mortality: End Stage Liver Disease and Liver Cancer from 2008–2016 In 2008, these annual analyses began tracking liver cancer and ESLD (cirrhosis) together, since in the California prison population both are sequelae of chronic viral hepatitis C infection. Chronic liver disease has consistently ranked as one of the top three causes of death in this population. In 2017, it accounted for 39 deaths, 10% of the total. Table 8 shows the observed numbers of deaths from chronic hepatitis C infection from 2008–2017. 15 Analysis of 2017 CCHCS Death Reviews TABLE 8. CCHCS CHRONIC HEPATITIS C ASSOCIATED DEATHS, 2008–2017. YEAR LIVER CANCER DEATHS CIRRHOSIS DEATHS TOTAL HEPATITIS C ASSOCIATED DEATHS CCHCS NUMBER OF INMATES CCHCS HEP C ASSOCIATED DEATH RATE PER 100,000 INMATES 2008 30 35 65 170,022 38.2 2009 30 60 90 169,459 53.1 2010 22 47 69 166,700 41.4 2011 23 53 76 161,843 47.0 2012 25 47 72 134,929 53.4 2013 27 43 70 133,297 52.5 2014 21 47 68 135,225 50.3 2015 19 37 56 128,824 43.5 2016 23 18 41 128,705 31.9 2017 18 21 39 130,807 29.8 Figure 4 shows the trended death rates from liver cancer, cirrhosis, and all chronic hepatitis C from 2008 to 2017. FIGURE 4. CCHCS CHRONIC HEPATITIS C ASSOCIATED DEATH RATES, 2008–2017. CCHCS Hep C Death Rate per 100,000 Inmates 60 45 30 15 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 In the past two years there appears to be a significant lowering of the death rate from hepatitis C sequelae which may coincide with the CCHCS practice of prescribing newly available treatments for chronic hepatitis C infection in appropriate candidates, as well as the emphasis on following the care guides for complications of hepatitis C. 16 Analysis of 2017 CCHCS Death Reviews D. Trends in Specific Causes of Mortality: Suicide In 2017, suicide was the sixth leading cause of death in the CCHCS. Table 9 and Figure 5 show numbers of deaths from suicide and the trended death rates by suicide for California state prisons compared to all US prisons. TABLE 9. NUMBERS AND RATES OF SUICIDE-RELATED DEATHS: CALIFORNIA, ALL U.S. STATE PRISONS, 2006–2017. YEAR CCHCS SUICIDES CCHCS SUICIDE RATE/100,000 U.S. STATE PRISON SUICIDE RATE/100,000 2006 43 25.1 17 2007 33 19.3 16 2008 38 22.3 15 2009 25 14.8 15 2010 34 20.4 16 2011 34 21 14 2012 32 23.7 16 2013 30 22.5 15 2014 23 17 20 2015 24 18.6 not available 2016 26 20.2 not available 2017 31 23.7 not available AVERAGE 31 20.7 16 FIGURE 5. SUICIDE DEATH RATES IN THE CALIFORNIA CORRECTIONAL SYSTEM (2006-2017) AND U.S. STATE PRISONS (2006–2014). CCHCS Suicide Rate/100,000 U.S. State Prison Suicide Rate/100,000 Linear (CCHCS Suicide Rate/100,000) Linear (U.S. State Prison Suicide Rate/100,000) 32.5 26. 25.1 22.3 20.4 19.3 19.5 21. 23.7 23.7 22.5 17. 14.8 18.6 20.2 13. 6.5 0. 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 17 Analysis of 2017 CCHCS Death Reviews Figure 5 indicates no significant trend in the rate for CCHCS or the rate for all U.S. State prisons, though the average suicide rate for California is 25% higher (20.7) than the average suicide rate for all U.S. state prisons through 2014 (16.0). E. Trends in Specific Causes of Mortality: Homicide In 2017, homicide was the seventh leading cause of death in the CCHCS. There were 19 deaths by homicide, representing 4.9% of all deaths. Table 10 shows numbers and rates of homicides in the California prisons (2006– 2017) and compares the rates with those of U.S. state prisons (2006 - 2014). TABLE 10. NUMBERS OF HOMICIDE-RELATED DEATHS IN CALIFORNIA AND ALL U.S. STATE PRISONS, 2006–2017. YEAR CCHCS HOMICIDES CCHCS HOMICIDE RATE/100,000 U.S. STATE PRISON HOMICIDE RATE/100,000 2006 16 9.3 4 2007 22 12.9 4 2008 7 4.1 3 2009 9 5.3 4 2010 23 13.8 5 2011 17 10.5 5 2012 21 15.6 7 2013 20 15 7 2014 9 6.7 7 2015 16 12.4 not available 2016 26 20.2 not available 2017 19 14.5 not available AVERAGE 17 11.7 5.1 Figure 6 graphs those same figures and shows the trends of homicide death rates over time. The homicide death rate in CCHCS prisons (11.7) is more than twice that of all U.S. state prisons (5.1). The rate of homicide deaths is increasing. In 2016, the death rate by homicide was the highest since 2006, but in 2017 the rate dropped closer to the historical average. Last year’s analysis mentioned a special report for the Bureau of Justice Statistics prepared in 2005, which noted that the homicide rate in all U.S. state prisons had dropped by 93% from 54/100,000 in 1980 to 4/100,000 in 2002. (bjs.gov/content/pub/pdf/shsplj.pdf) 18 Analysis of 2017 CCHCS Death Reviews FIGURE 6. HOMICIDE DEATH RATES IN THE CALIFORNIA CORRECTIONAL SYSTEM AND ALL U.S. STATE PRISONS, 2006–2017. CCHCS Homicide Rate/100,000 Linear (CCHCS Homicide Rate/100,000) U.S. State Prison Homicide Rate/100,000 Linear (U.S. State Prison Homicide Rate/100,000) 25 20.2 20 15 10 15.6 13.8 12.9 9.3 5 15. 14.5 12.4 10.5 6.7 4.1 5.3 2008 2009 0 2006 2007 2010 2011 2012 2013 2014 2015 2016 2017 F. Trends in All Preventable Deaths, 2006–2017 The rates of all (definitely) preventable and possibly preventable deaths are shown for each year from 2006 to 2017 in Table 11. In 2017, that rate was 11.5/100,000. TABLE 11. RATES OF PREVENTABLE DEATHS AMONG CALIFORNIA INMATES, 2006–2017. PREVENTABLE DEATHS YEAR DEFINITELY POSSIBLY ALL INMATE POPULATION ALL PREVENTABLE DEATH RATE PER 100,000 INMATES 2006 18 48 66 total 171,310 38.5 2007 3 65 68 total 170,786 39.8 2008 5 61 66 total 170,022 38.8 2009 3 43 46 total 169,459 27.1 2010 5 47 52 total 166,700 31.2 2011 2 41 43 total 161,843 26.6 2012* 1 42 43 total 134,929 31.9 2013 0 35 35 total 133,297 26.3 2014 0 24 24 total 135,225 17.7 2015 0 12 12 total 128,824 9.3 2016 0 18 18 total 128,705 14.0 2017 1 14 15 total 130,807 11.5 *Note: Figures for 2012 here are as reported in the original Analysis of 2012 Death Reviews. An error in the 2012 data was subsequently introduced into this table in the Analysis of 2014 Death Reviews. It has been corrected here. 19 Analysis of 2017 CCHCS Death Reviews Figures 7 and 8 show the favorable downward trend in overall preventable death which began in 2009, the fourth year of the Receivership. FIGURE 7. NUMBER OF PREVENTABLE DEATHS IN THE CALIFORNIA CORRECTIONAL HEALTHCARE SYSTEM, 2006-2017. Possibly Preventable 80 60 3 20 0 5 18 40 48 2006 65 2007 Preventable 61 2008 3 5 43 47 2009 2 2010 1 34 42 2011 2012 0 0 35 2013 24 2014 0 1 12 18 14 2015 2016 2017 0 FIGURE 8. PREVENTABLE DEATH RATES IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017. Preventable Death Rate Per 100,000 Inmates 50 40 Linear ( Preventable Death Rate Per 100,000 Inmates ) 38.5 39.8 38.8 27.1 30 31.2 26.6 31.9 26.3 17.7 20 9.3 10 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 14. 11.5 2016 2017 There appear to be three periods reflected in Figure 8. The first three years of the receivership, 2006–2008, was a period during which a major goal was the identification and elimination of unsafe practicing physicians. During this period, the overall preventable death rate averaged 39.0. From 2009–2013 there was a period of significant improvement. This period coincides with the Receiver’s redesign of the system of care. The Receiver’s Turnaround Plan submitted in 2008 emphasized timely access to competent medical providers in a system of primary care, timely access to prescribed medication and treatment, timely access to a system of specialty care, and to the construction of new healthcare infrastructure including a new medical prison facility intended to house the chronically medically ill. During this five-year period, the overall preventable death rate averaged 28.6. The last four years, 2014–2017, coincides with the intent to create a culture of quality improvement and a further maturation of the redesign described above, and the opening of the California Health Care Facility. In addition, a mandated reduction in the prison population intended to reduce the severe overcrowding in the prisons, was successful in reducing the total CCHCS population from an average of 168,353 (2006–2011) to an average of 131,965 (2012–2017), a reduction of 21.6%. During this most recent four-year period, the overall preventable death rate averaged 13.1. 20 Analysis of 2017 CCHCS Death Reviews FIGURE 8A. THREE PERIODS IN PREVENTABILITY TRENDS, CCHCS 2006-2017. Preventable Death Rates per 100,000 Inmates 50 40 38.5 2006-08 2009-13 2014-17 Linear ( 2006-08 ) Linear (2009-13) Linear (2014-17) 39.8 38.8 27.1 30 31.2 31.9 26.6 26.3 17.7 20 9.3 14. 11.5 2016 2017 10 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 G. Trends in Care Lapses 1. The Relationship Between the Number of Lapses and Patient Complexity Lapses occur frequently in the practice of medicine in every system of care, but because patients are basically healthy, the majority of lapses in an outpatient setting do not result in significant adverse clinical outcomes. When patients are sicker and more complicated, with more chronic medical conditions coexisting with severe mental illness or substance abuse, they require more prescription medications, more specialty care, more emergency department visits and hospitalizations. Sicker patients are at greater risk of experiencing an adverse outcome. In 2015 and 2016, this review looked at all of the patients who died and counted the number of chronic medications prescribed and the number of associated medical conditions exclusive of the primary cause of death. Table 12 shows the findings in those two years. TABLE 12. FREQUENCY OF ASSOCIATED CONDITIONS (EXCLUSIVE OF PRIMARY CAUSE OF DEATH) IN CCHCS INMATE DEATHS, 2015 AND 2016. CONDITION NUMBER OF CASES 2015 2016 Hypertension (HTN) 170 145 Hepatitis C 114 88 Diabetes mellitus (DM) 87 78 Dyslipidemia (DLP) 52 68 Severe mental illness 70 67 Coronary artery disease (CAD) 62 56 21 Analysis of 2017 CCHCS Death Reviews CONDITION NUMBER OF CASES 2015 2016 Chronic obstructive pulmonary disease (COPD) 54 53 Gastroesophageal reflux disorder (GERD) 27 38 Chronic pain/Osteoarthritis * 27 Chronic kidney disease (CKD) * 26 Cancer * 26 Asthma 21 26 Benign prostate hypertrophy (BPH) 36 25 Obesity * 25 Cancer-liver, ESLD * 24 Seizure disorder 17 23 Congestive heart failure 18 * Atrial fibrillation 16 * Coccidioidomycosis 15 * History of stroke (Cerebrovascular accident) 13 * Hypothyroidism 13 * Other Conditions (appearing in fewer than 10 cases each, including those marked ‘*’ above) 176 231 TOTAL ASSOCIATED CONDITIONS 961 1155 Total Deaths 355 334 AVERAGE ASSOCIATED CONDITIONS PER DECEDENT 2.7 3.5 9 0-37 9 0-24 PRESCRIBED MEDICATIONS PER DECEDENT: AVERAGE RANGE Prescribed Medications – The 334 decedents in 2016 were taking an average of nine prescription medications (range zero to 24). The 355 patients who died in 2015 were also prescribed an average of nine medications (range zero to 37). Associated conditions – The average number of associated conditions for the 355 decedents in 2015 was 2.7 and for the 334 decedents in 2016 was 3.5. These conditions are in addition to the designated cause of death. Their contribution to the overall burden of chronic disease within the primary care setting adds to the complexity of management and increases the chance for care lapses to occur. 22 Analysis of 2017 CCHCS Death Reviews There is no reason to believe that the decedents in 2017 were any less complex and any less susceptible to a higher incidence of care lapses. 2. The Relationship Between Number of Lapses and Preventability of Death in 2017 Prior annual death report analyses have shown a relationship between the number of lapses occurring in a single case and a cascade of consequences which can lead to preventable death. In the medical literature, this has been called the “Swiss cheese effect” – multiple errors lining up to result in an adverse outcome. The findings for 2017 reinforce this observation. Table 13 shows that in 2017 the average number of lapses in possibly preventable deaths (1.2) were four times the average number of lapses in the not preventable deaths (0.3). There were two lapses contributing to the one definitely preventable death in 2017. TABLE 13. NUMBER OF LAPSES BY CATEGORY OF PREVENTABILITY, 2017. PREVENTABILITY # DEATHS # LAPSES AVERAGE LAPSES/DEATH 1 2 2.0 14 17 1.2 373 108 0.3 Preventable Possibly preventable Not preventable FIGURE 9. AVERAGE NUMBER OF LAPSES PER CASE BY PREVENTABILITY, 2007–2017. Likely Preventable 7. Non-Preventable 6.2 5.25 4.4 3.7 3.7 3.5 1.75 Possibly Preventable 2.4 1.7 0.5 0.6 3.1 2.2 0.6 0.8 3. 2.5 0.5 2. 2.8 2.6 0.6 0. 0.6 2.2 0. 0.4 1.9 1.6 0. 0.5 0. 0.5 2. 1.2 0.3 0. 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Figure 9 shows trended data for the average number of lapses per case by their preventability determination. For all years, lapses in (definitely) preventable cases averaged 3.8, in possibly preventable cases averaged 2.3 and in not preventable cases averaged 0.5. Table 14 shows the total number of lapses from 2007 to 2017. 23 Analysis of 2017 CCHCS Death Reviews TABLE 14. NUMBER OF LAPSES, BY PREVENTABILITY, IN CCHCS DEATHS, 2007–2017. YEAR DEFINITELY PREVENTABLE POSSIBLY PREVENTABLE NOT PREVENTABLE TOTAL NO. OF LAPSES NO. OF CASES AVG LAPSES PER CASE # % # % # % 2007 11 4% 109 36% 179 60% 299 395 0.8 2008 22 6% 147 41% 193 53% 362 369 1.0 2009 11 4% 90 29% 205 67% 306 393 0.8 2010 31 7% 147 32% 284 61% 462 415 1.1 2011 6 2% 92 37% 154 61% 252 388 0.6 2012 2 1% 105 34% 198 65% 305 362 0.8 2013 0 0% 97 32% 206 68% 303 366 0.8 2014 0 0% 53 31% 120 69% 173 319 0.5 2015 0 0% 19 10% 176 90% 195 355 0.5 2016 0 0% 35 18% 163 82% 198 334 0.6 2017 2 2% 17 13% 108 85% 127 388 0.3 FIGURE 10. TREND IN ANNUAL AVERAGE OF CARE LAPSES PER DEATH, CCHCS, 2007–2017. Average Lapses per Case Linear (Average Lapses per Case) 1.5 1.0 0.5 0.0 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 The last column in Table 13 is trended in Figure 10, showing that the annual number of care lapses for all cases has been trending downward since 2010, and in 2017 the rate was 0.3 lapses per case – the lowest in the history of the Receivership. 24 Analysis of 2017 CCHCS Death Reviews VIII. TARGETED OPPORTUNITIES FOR IMPROVEMENT A. The Primary Care Model and Preventable Deaths Planning for a primary care model of care in the CCHCS began in 2007, and by 2009 it had been partially implemented in all California state prisons. A primary care medical home creates an environment for ensuring continuous, integrated, coordinated and planned care, especially for patients with chronic or complex combinations of medical illness. Primary care teams are expected to have accountability for patient outcomes, to advocate on behalf of their patients and to use evidence-based guidelines in managing chronic conditions. They are responsible for timely access to appropriate care including specialty referrals, and for coordinating follow-up care after their patients are sent to hospital emergency rooms or experience hospitalizations. TABLE 15. IDENTIFIABLE PRIMARY CARE IN CALIFORNIA INMATE DEATH CASES, 2009–2017. YEAR CASES WITH IDENTIFIED PRIMARY CARE PHYSICIAN TOTAL DEATHS % OF TOTAL PREVENTABLE DEATH RATE PER 100,000 INMATES 2006 not available 424 NA 38.5 2007 not available 395 NA 39.8 2008 not available 369 NA 38.8 2009 141 393 35.5% 27.1 2010 217 415 52.3% 31.2 2011 209 388 53.4% 26.6 2012 230 367 62.7% 30.4 2013 240 366 65.6% 26.3 2014 200 319 62.7% 17.7 2015 237 355 66.8% 9.3 2016 235 334 70.4% 14.0 2017 207 388 54.9% 11.5 In 2009, the DRC began identifying patients who had an identifiable primary care physician (PCP) and were also looking for instances in which the primary care model was not working well. Care lapses types 5 and 6 directly address the primary care model. Table 15 shows, for 2009–2017, the number and percentage of cases in which the reviewer could identify a primary care clinician. As shown in Figure 11, there appears to be an inverse correlation between the percent of decedents with an identified PCP and the overall preventable death rate. 25 Analysis of 2017 CCHCS Death Reviews FIGURE 11. PERCENTAGE OF DEATHS IN THE CCHCS WITH AN IDENTIFIED PRIMARY CARE PHYSICIAN, AND CORRESPONDING RATES OF PREVENTABLE DEATH, 2009–2017. % Decedents with PCP Preventable death rate per 100,000 inmates 80% 50 60% 40 30 40% 20 20% 10 0% 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 B. Trends in Specific Targeted Causes for Preventable Death TABLE 16. NUMBERS AND RATES OF PREVENTABLE DEATHS FROM CARDIOVASCULAR, END STAGE LIVER DISEASE, AND CANCER IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017. YEAR PREVENTABLE CARDIOVASCULAR DEATHS Number Rate/100,000 PREVENTABLE ESLD AND LIVER CANCER DEATHS PREVENTABLE (NON-LIVER) CANCER DEATHS Number Rate/100,000 Number Rate/100,000 2006 18 10.5 2 1.2 6 3.5 2007 16 9.4 6 3.5 7 4.1 2008 14 8.2 4 2.4 9 5.3 2009 9 5.3 4 2.4 10 5.9 2010 7 4.2 2 1.2 4 2.4 2011 11 6.8 1 0.6 6 3.7 2012 8 5.9 3 2.2 1 0.7 2013 7 5.3 4 3.0 4 3.0 2014 10 7.4 2 1.5 6 4.4 2015 3 2.3 1 0.8 1 0.8 2016 2 1.6 5 3.9 2 1.6 2017 5 3.8 0 0.0 3 2.3 26 Analysis of 2017 CCHCS Death Reviews Table 16 shows the raw data from which the run charts are constructed for three specific types of preventable death: 1. cardiovascular, 2. end stage liver/ liver cancer and 3. (non liver) cancer. 1. Preventable Cardiovascular Death Rates FIGURE 12. PREVENTABLE CARDIOVASCULAR DEATHS – NUMBER OF PREVENTABLE CASES AND RATES OF DEATH IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017. rate per 100000 Preventable Cardiovascular Deaths Linear (rate per 100000) 20 15 10 5 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Figure 12 demonstrates a continuing reduction in preventable deaths from cardiovascular (CV) disease, attributable to the CCHCS emphasis on better recognition and management of red flag symptoms and signs of heart attack (reduction of type 1 lapses) and on better management of acute and chronic heart disease syndromes and evidence-based treatment of CV risk factors. All of these are addressed in the CCHCS Care Guidelines for Chest Pain, Hypertension, Dyslipidemia and Diabetes Mellitus. Failures to manage patients according to these Care Guidelines may result in a type 2 care lapse. 2. Preventable End Stage Liver Disease (including liver cancer) Death Rates Figure 13 shows the run chart for the number of preventable cases and rates of death from end stage liver disease, including hepatocellular carcinoma (liver cancer). FIGURE 13. PREVENTABLE END STAGE LIVER DISEASE DEATHS – NUMBER OF PREVENTABLE CASES AND RATES OF DEATHS IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017. rate per 100000 20 Preventable ESLD deaths Linear (rate per 100000) 15 10 5 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 In 2016, there were five cases of preventable death because of failure to follow the guideline for ultra sound screening intended to detect early treatable hepatocellular cancer in patients with chronic hepatitis C. The DRC recommended a targeted educational effort directed at the care teams in the CCHCS. And in 2017 there were no cases of preventable deaths due to end stage liver disease and liver cancer. 27 Analysis of 2017 CCHCS Death Reviews 3. Preventable (Non-liver) Cancer Death Rates As seen in Figure 14, the improvement in preventable cancer deaths continues, especially during the past three years. In 2017, the three possibly preventable deaths from cancer were all attributed to delays in diagnosis. Two of these deaths resulted from failure to evaluate “red flag” symptoms. Weight loss signaled a cancer of the stomach and diagnosis was delayed by 8 months. An abnormal testicular examination indicated a testicular cancer, but there was a diagnostic delay of 13 months. And an abnormal laboratory test showing anemia and a positive test for blood in the stool signaled a colon cancer but there was a delayed diagnosis of 18 months. FIGURE 14. PREVENTABLE CANCER DEATHS – NUMBER OF PREVENTABLE CASES AND RATES OF DEATH IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017. rate per 100000 Preventable Cancer Deaths Linear (rate per 100000) 20 15 10 5 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 4. Deaths from Drug Overdose Table 17 shows CCHCS annual rates of death from drug overdose from 2006 to 2017. Similar rates for all US prisons are also shown, but data is only available through 2014. TABLE 17. NUMBERS AND RATES OF DRUG OVERDOSE-RELATED DEATHS IN THE CALIFORNIA CORRECTIONAL HEALTHCARE SYSTEM AND IN ALL U.S. PRISONS, 2006–2017. YEAR CCHCS DRUG OVERDOSES CCHCS RATE/100,000 U.S. STATE PRISON RATE/100,000 2006 17 9.9 4 2007 9 5.3 3 2008 19 11.2 4 2009 14 8.3 4 2010 23 13.8 3 2011 12 7.4 4 2012 15 11.1 3 2013 24 18.0 4 2014 19 14.1 4 2015 19 14.7 not available 2016 29 22.5 not available 2017 40 30.6 not available Avg 20 13.9 3.7 28 Analysis of 2017 CCHCS Death Reviews FIGURE 15. DRUG OVERDOSE DEATH RATES IN THE CALIFORNIA CORRECTIONAL HEALTHCARE SYSTEM, 2006–2017. CCHCS Overdose Rate/100,000 CCHCS drug overdoses Linear (CCHCS Overdose Rate/100,000 ) 50 40 30 20 10 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Death from drug overdose in the CHCS continues to rise at a very significant rate, mirroring the experience in American society at large. For the years 2006–2014, the CCHCS drug overdose death rate averaged 12.6, a little more than three times the total US prison rate of 3.7. For the years 2015–2017, the rate has averaged 22.6, and has been increasing every year. There are no comparable rates for all US prisons in that period. In 2017, there were 40 cases of drug overdose. Only one case involved prescribed medication, and the drugs responsible in that case were a combination of antipsychotic agents. As seen in Table 1, all opioid overdoses were due to non prescribed illicit sources of heroin or other opioids. The most common illicit agents used in these cases were heroin or morphine (22), fentanyl (9), and methamphetamines (13). These numbers add up to more than 40 because in 6 cases there were combinations of drugs used. The opioid antagonist Narcan was utilized unsuccessfully during the emergency resuscitation in 35 of the 39 cases involving narcotic agents. In most of these cases, the patients had been unresponsive for many minutes to hours before staff were notified and emergency protocols instituted. The CCHCS prisons are not shielded from the "opioid epidemic”. Illicit opioids and amphetamines are widely available and patients are especially vulnerable to their effects, including death by overdose. 5. Coccidioidomycosis Death Rates Coccidioidomycosis, a fungal infection endemic to the eight prisons located in the San Joaquin corridor of central California, was responsible for as many as 9 deaths annually in susceptible inmates. A Federal order to remove high risk patients from these 8 prisons has resulted in a major success story for the CCHCS. Compared to the 47 deaths from cocci from 2006 to 2013, there were only two deaths from 2014 to 2016, and none in 2017. 29 Analysis of 2017 CCHCS Death Reviews TABLE 18. COCCIDIOIDOMYCOSIS RELATED DEATHS IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017. YEAR COCCI RELATED DEATHS 2006 9 2007 6 2008 6 2009 5 2010 7 2011 3 2012 7 2013 4 2014 0 2015 1 2016 1 2017 0 Figure 16 is a run chart trending numbers of deaths and death rates from coccidioidomycosis. FIGURE 16. COCCIDIOIDOMYCOSIS RELATED DEATHS AND DEATH RATES IN THE CALIFORNIA CORRECTIONAL SYSTEM, 2006–2017. rate per 100000 10 8 6 4 2 0 2006 2007 2008 Cocci related deaths 2009 2010 2011 2012 Linear (rate per 100000) 2013 2014 2015 2016 2017 30 Analysis of 2017 CCHCS Death Reviews IX. PERFORMANCE IMPROVEMENT PLANS AND TOOLS In past annual analyses the different performance improvement strategies, plans, and tools developed and utilized by the receivership have been described in some detail. These have included, but are not limited to, the following: A. The CCHCS Statewide Performance Improvement Plan 2016–2018 This plan (https://cchcs.ca.gov/wp-content/uploads/sites/60/2017/08/T31_20160201_Appendix1.pdf) highlighted the adoption of The Complete Care Model, emphasizing continuous, coordinated, comprehensive and planned patient centered care with a focus on access, prevention and population health management. Population health management strategies include the use of registries for patients with chronic conditions, and the distribution of indicator dashboards to track the performance of health care teams in meeting standards of care. For example, the end stage liver disease registry dashboard tracks the percentage of patients who have received liver cancer screening ultrasounds, as well as the prescription of certain medications known to be of value in preventing complications of severe liver disease. The Health Care Services Dashboards can be accessed here: https://cchcs.ca.gov/wpcontent/uploads/sites/60/2018/10/Public-Dashboard-2018-08.pdf. There are dashboards common to all of the state prisons, distributed monthly. These are useful for monitoring a host of key performance indicators in the management of chronic diseases like asthma, diabetes and advanced liver disease. Patients on ten or more chronic medications are subject to medication reconciliation. Adherence to scheduling and access standards for primary care and specialty care are monitored. Routine screening for colon cancer and women’s health maintenance are tracked. These monthly dashboards are used by health care managers in each of the 35 CCHCS facilities to track performance and to target areas for improvement. B. CCHCS Care Guides The Care Guides (https://cchcs.ca.gov/clinical-resources/) are tools for use by clinicians and care teams in the management of patients with the following conditions: Anticoagulation, Asthma, Chest Pain, Chronic Wound Management, Clozapine, Coccidioidomycosis, Chronic Obstructive Pulmonary Disease, Cognitive Impairment/Dementia, Diabetes, Dyslipidemia (high or abnormal cholesterol), End Stage Liver Disease, Gender Dysphoria, Hepatitis C, HIV, Hunger Strike (fasting and referring), Hypertension, Major Depressive Disorder, Pain Management, Palliative Care, Schizophrenia, Seizure Disorders, Skin and Soft Tissue Infections, and Tuberculosis. Similar resources for nurses are also in use and include Protocols and Encounter forms for patients with Abdominal Trauma, Allergic Reaction(s), Asthma, Burns, Chest Pain, Chest Trauma, Constipation, Dental Conditions, Earache, Epistaxis, Eye injury/ irritation, Female Genitourinary Complaints, Headache, Hemorrhoids, Rash, Insect Stings, Intravenous Therapy, Loss of Consciousness, (non traumatic) Musculoskeletal Complaints, Respiratory Distress, Seizure, Tetanus Prophylaxis, Upper Respiratory Infections, and Wound Care. 31 Analysis of 2017 CCHCS Death Reviews C. The Electronic Medical Record Statewide implementation of the Electronic Health Records System was completed in November 2017. X. CONCLUSIONS This twelfth annual review shows a continuing increase in the all cause CCHCS death rate, possibly due to an increase in the average age of the California prison inmate population cause by release of large numbers of younger inmates. In 2017, the overall death rate of 297/100,000 was the highest in the past 12 years. Although most of this increase can be attributed to non preventable deaths from cancer and cardiovascular disease, there has also been a continued dramatic rise in deaths by illicit opioid and amphetamine drug overdose. The CCHCS under the Federal Receiver has continued its major redesign of the system of care, which has resulted in significant improvement in major outcomes, achieved for a complex patient population. There was a significant decline in the number of cited care lapses in 2017, with 0.3 lapses per case the lowest rate in the eleven years that this measure has been tracked. The death rate from preventable cardiovascular deaths continued a favorable downward trend. A targeted effort to improve the monitored surveillance for liver cancer in end stage liver disease resulted in a significant reduction in possibly preventable liver cancer deaths, which went from 5 cases in 2016 to 0 cases in 2017. Overall preventable death rates have continued to decrease. The fifteen total preventable deaths in 2017 represented a rate of 11.5/100,000. This compares with the average rate of 39/100,000 during the first three years of the Receivership, and continues the favorable downward trend. The success of the Receivership in transforming healthcare in the California state prisons has resulted in a process of revocable delegation. By March of 2018, 16 of the 35 California prisons had been delegated from the Receivership back to the State of California based on favorable reviews of medical care by the Office of the Inspector General, and subject to ongoing periodic monitoring by the Receiver. These and other successes can be followed in the triennial reports of the Receiver to the Federal Judiciary: https://cchcs.ca.gov/wp-content/uploads/sites/60/2018/06/T38_20180601_TriAnnualReport.pdf 32