CLN - The Sample

Drug Overdoses Follow Exponential 40-year Upward Growth Rate

The Sample: November 2018

The annual mortality rate from unintentional overdoses of all drug classes has followed a “remarkably smooth mathematical trajectory” since 1979, predating the current opioid epidemic, and suggesting that “prevention efforts must extend beyond control of specific drugs to address deeper factors driving the epidemic,” according to the authors of a new analysis (Science 2018;361:eaau1184).

Even with this overall increased rate, mortality patterns for individual drugs rose and fell over the years with “no regular or predictable pattern” for any.

The investigators sought to understand the opioid epidemic in the context of the “complex and evolving dynamics of drug use” in the U.S. To do so they examined mortality patterns of all nearly 600,000 accidental drug poisonings reported from 1979 through 2016 through the U.S. National Vital Statistics System. In this nearly 40-year period, they found the overall mortality rate per 100,000 rose from 1.13 in 1979 to 16.96 in 2016.

Embedded within this overall rise, the researchers found drug-specific subepidemics over time and in different areas of the country. For example, heroin deaths shot up starting in 2010, and heroin hot spots changed over time from being tightly clustered around major metropolitan areas to being more diffused, particularly in the Northeast and Southwest. The only region without notable drug hot spots is the upper Midwest (Northcentral).

The researchers also found distinct overdose patterns based on age, with individuals ages 20-40 more likely to have died of heroin and synthetic opioids versus those ages 40-60 more likely to have died from taking prescription opioids and other unspecified drugs.

The authors speculated that economic and technologic “push” factors such as more efficient communications and drug manufacturing as well as social and psychological “pull” factors like dissolution of communities and loss of purpose might be behind the trends they observed.

EHR-Based Diagnostic Protocol Speeds Assessment, Safe Discharge of Suspected Acute Coronary Syndrome Patients

An accelerated electronic health record (EHR)-based diagnostic protocol incorporating two cardiac troponin (cTn) measurements improved identification of emergency patients with chest pain who could be sent home safely without undergoing comprehensive cardiac testing or hospitalization (Circ Cardiovasc Qual Outcomes 2018; doi: 10.1161/CIRCULATIONAHA.118.036528).

Of the 31% of patients the pathway classified as low risk, just 0.4% died or had a myocardial infarction (MI) within 30 days. In comparison to normal care, the pathway also identified more patients as having MI, with an adjusted odds ratio of 1.36, and reduced by 6% hospitalizations within 30 days, with an adjusted odds ratio of 0.79.

“Given its ability to reduce the utilization of healthcare resources, the protocol may provide a model for health systems to provide safe and high-value care to emergency room patients with chest pain at lower cost,” wrote the investigators.

The HEART Pathway yields a score based on five inputs, including patient history, electrocardiogram (ECG) reading, age, risk factors, and two cTn measurements taken at 0 and 3 hours of a patient’s assessment. Patients with scores of 3 or lower on the first four inputs (identified collectively as HEAR) and without elevated cTn results were recommended for emergency department (ED) discharge without more extensive cardiac testing. Those with HEAR scores ≥4, elevated cTn levels, known cardiac artery disease, or ischemic ECG changes were designated as “non-low-risk” and in need of additional testing and/or hospitalization.

The pre/post prospective study involved 8,474 adult ED patients undergoing assessment for possible acute coronary syndrome at three Wake Forest Baptist Health hospitals in North Carolina. The investigators evaluated the 30-day hospitalization rate and secondary outcomes including objective cardiac testing, early discharge rates, index visit length of stay (LOS), and ED LOS. They considered these outcomes in cohorts of 3,713 and 4,761 patients seen in the 12 months before and after HEART Pathway was implemented, respectively.

Once HEART Pathway went live, providers would see an EHR popup advising that this tool was a best practice advisory. They would then be prompted to answer a series of questions to prospectively risk-stratify patients. cTn test results were linked directly from the laboratory information system.

Because most ED providers were aware of and informally using HEART Pathway before the tool went live, the authors suggested that their findings might reflect “contamination” that “may have decreased the effect size of our intervention,” suggesting that institutions that newly introduce the tool could realize more dramatic improvements.

29-biomarker Signature Promising Early Diagnostic for Pancreatic Ductal Adenocarcinoma

A proteomic analysis involving more than 1,700 case-control samples developed and validated a 29-analyte biomarker signature as an early-stage indicator of pancreatic ductal adenocarcinoma (PDAC) (J Clin Oncol 2018;36: If these findings are supported in a prospective validation study, the signature might be used as an early diagnostic tool for PDAC, according to the authors.

The investigators developed the biomarker signature from a cohort of 443 patients with PDAC (16 stage I, 132 stage II, 65 stage III, and 230 with stage IV) and 888 controls. The researchers used a backward elimination algorithm in each training set used to interrogate the candidate biomarkers, excluding one at a time if it did not improve the classification.

The final signature includes only the highest-ranked biomarkers, incorporating analytes such as apolipoprotein A1, complement C3, C4, and C5, intercellular adhesions molecule 1, and Lewis x, most involved in immunoregulatory processes. The signature yielded an area under the receiver operator characteristic (AUROC) of 0.91, 1.0, 0.99, 0.98, 0.99, and 0.98 for no cancer versus PDAC stages IA, IB, IIA, IIB, III, and IV.

The researchers validated the signature in a separate case-control cohort of samples from 219 controls and 443 patients with PDAC (15 stage I, 75 stage II, 15 stage III, and 38 stage IV). In the validation group the overall AUROC was 0.963, correlating to a specificity and sensitivity of 95% and 93% for stage I and II.