NAS cases identified among neonates with select diagnosis codes in any position:. FY 2008-2015: ICD-9 CM code 7795 (newb
Neonatal Abstinence Syndrome in Missouri 2008-2017
1
Neonatal Abstinence Syndrome (NAS)
Postnatal drug withdrawal syndrome most prevalent in opioid-exposed infants shortly after birth
Manifests in central nervous irritability, GI tract dysfunction and overactive autonomic response
Limited evidence on long-term effects
National rates of NAS increased by 300% between 1999 and 2013 –
1.5 per 1,000 births in 1999
–
6.0 per 1,000 births in 2013
Ko JY, Patrick SW, et al. (2016, August). Incidence of Neonatal Abstinence Syndrome—28 States, 1999-2013. MMWR 16(31)
NAS cases identified among neonates with select diagnosis codes in any position:
FY 2008-2015: ICD-9 CM code 7795 (newborn drug withdrawal syndrome)
FY 2016-2017: ICD-10 CM code P961 (neonatal withdrawal symptoms from maternal use of drugs) and P962 (Withdrawal symptoms from therapeutic use of drugs in newborn)
Source: Hospital Industry Data Institute, 2008-2017 Missouri Hospital Inpatient and Outpatient Discharge Databases
2
9
400%
8
350%
7
300%
6
250%
5 200% 4 150%
3
100%
2 1
50%
0
0% 2008
2009
2010
2011
2012
NAS Rate per 1,000 Births
2013
2014
2015
2016
NAS Percent Change from 2008
Number of NAS Diagnoses per 1,000 Births
Missouri NAS Rate Identified with Conventional Hospital Discharge Coding Surveillance: 2008-2017
2017
NAS Rate % Change from 2008 3
Characteristics of Neonates with NAS in Missouri: 2016-2017 1,083 Median Length of Stay
Non-NAS 2 Days
Newborns with NAS remain in the hospital on average 10 days longer than newborns without NAS
Hospital charges for newborns with NAS are 6.7 times the average amount for newborns without NAS
Medicaid is the expected primary payer for more than 8 out of every 10 NAS births in Missouri
NAS 12 Days Median Charges
Non-NAS $4,894 NAS $32,938
Primary Payer
9% 8% 50% 6%
83%
44%
NAS Medicaid
Uninsured
Non-NAS Commercial/Other
Diagnostic Related Group
Non-NAS
NAS 0%
20%
40%
60%
80%
100%
With major problems Died or transferred Extreme immaturity/resp. distress Normal newborn Other significant problems
4
Potential Under-Coding of NAS
5
Anecdotal Evidence and Opioid-Related Claims Data Linked to New and Expectant Mothers Suggest Under-coding of NAS 1,083 Neonates diagnosed with NAS at birth in Missouri during fiscal years 2016 and 2017
Type of Opioid Diagnosis for New and Expectant Mothers in Missouri: 2016-2017
3,714 Unique new and
expectant mothers with a hospital encounter for opioids during same time period:
Average of 5.5 opioid-related visits per patient For women who gave birth during period, 56% of opioid visits occurred during gestation (62% occurred during the third trimester)
18% 27%
9%
10% 21% 15% Opioid dependence, uncomplicated Long term (current) use of opiate analgesic Opioid abuse, uncomplicated Opioid dependence with withdrawal Opioid use, unspecified, uncomplicated Other opioid diagnosis 6
Number of Opioid-Related Hospital Visits by Unique New and Expectant Mothers in Missouri in Relation to Date of Birth: 2016-2017 ← Prior to Gestation: 1,016 Eventual Mothers
3rd Trimester & Day of Birth: 2,564 Expectant Mothers
1,600 1,400 1,200 1,000 800 600 400
1st Trimester: 589 Expectant Mothers
2nd Trimester: 839 Expectant Mothers
200 0
Number of Opioid Visits
Number of Unique Mothers 7
Incidence of NAS in Missouri by County During 2016 and 2017: Rate per 1,000 Births Identified with Diagnosis Codes for the Infant vs. Linking New and Expectant Mothers to Hospitalizations for Opioid Misuse Rate of NAS Detected by Diagnosis on Hospital Discharge Record, 2016-2017
Rate of New and Expectant Mothers with an Opioid-Related Inpatient or ED Visit: Rate per 1,000 Births by County, 2016-17
8
2018 Survey of Missouri Obstetric Hospitals on the Frequency & Severity of NAS
9
Survey to Missouri Hospital Obstetric Departments on Perceived Severity of NAS, Screening Protocol, Accuracy of Coding and Barriers to Reducing the Incidence of NAS
Survey administered to 70 birthing hospitals during April, 2018:
Responding hospitals accounted for 33,785 total and 257 NAS births for Missouri newborns during 2017:
41 responses received from 38 hospitals (54% response rate)
45% of all births and 50% of all NAS births in Missouri
Key Findings:
Perceived severity of NAS in respondents’ hospitals featured agreement with actual NAS rates identified through claims data
Large differences observed between the estimated frequency of NAS births at respondent hospitals and actual NAS births identified with claims data
–
2,369 survey-estimated vs. 257 NAS births identified with claims data at responding hospitals during 2017
–
Claims-based NAS rate per 1,000 births = 7.6
–
Survey-estimated NAS rate per 1,000 births = 70.1
Perceived accuracy of claims-based NAS coding in respondents’ hospitals featured disagreement with differences between survey-estimated and claims-identified NAS births
10
On a scale of 1 to 10 (with 10 being most severe), how would you describe the frequency of newborns with NAS delivered in your hospital? Distribution of Survey Respondents by Reported NAS Severity, Estimated NAS Percentage and Actual NAS Percentage Identified by Hospital Claims During 2017 14% 12% 12
12%
10
10%
8
8% 6%
6
6%
6% 4%
4 2 0
Percent of Births with NAS
Number of Responding Hospitals
14
1% 0.4% 1
0.3%
1.2%
0.9%
0% 2 to 3 4 to 6 Survey Reported NAS Severity (1-10 Scale)
# Hospitals
2%
Survey Reported NAS %
8 to 9
Claims Identified NAS % 11
Survey-Estimated NAS Births and Actual NAS Births Identified by Hospital Claims During 2017 by Reported NAS Severity 1,200
16,000
1,111 1,012
1,000
14,000
778
800
10,000
657
8,000
600 396
400
6,000
369
Number of Total Births
Number of NAS Births
12,000
4,000 200
0
121 10
37 27
2,000
99
27 0
1
2 to 3 4 to 6 Survey Reported NAS Severity (1-10 Scale)
Claims Identified NAS Births
Survey Estimated NAS Births
8 to 9
Difference
All Births 12
On a scale of 1 to 10 (with 10 being most accurate), how would you describe the accuracy of diagnosing newborns with NAS at your hospital (i.e. presence of ICD-10 codes P961 or P962 on the discharge record)? Survey Estimated NAS Births and Actual NAS Births Identified by Hospital Claims During 2017 by Reported NAS Coding Accuracy 8% 7% 20
6% 5%
15
4% 10
3% 2%
5
Percent of Births with NAS
Number of Responding Hospitals
25
1% 0
0% 1 to 3
4 to 7 8 to 10 Survey Reported NAS Claims Coding Accuracy (1-10 Scale)
# Hospitals
Survey Reported NAS %
Claims Identified NAS % 13
Does your hospital use an evidence-based assessment tool that allows for scoring of an infant with signs and symptoms of NAS to base further treatment on? If yes, which tool?
Finnegan 42%
Yes 94% No 6% Modified Finnegan 22%
NAS Score 14% Eat, Sleep, Console 6% Eat, Sleep, Console & Other 6% Neonatal Withdrawal Inventory 6%
14
Does your hospital screen maternal patients for opioid and other substance use disorders upon admission and/or triage? If yes, how?
Patient History/Risk Behavior 24%
Yes 84%
Verbal Assessment 22%
No 16%
Urine Drug Screen Based on Criteria 16%
Other 14%
Questionnairre 8%
15
Word Cloud of Responses to: Describe the top three barriers to reducing maternal OUD/SUD and NAS in your community (N=37)
16
Conclusion
Cases of NAS are growing more rapidly than the broader opioid epidemic
In Missouri, rates of hospital inpatient and emergency department utilization for opioid misuse across all age groups has increased by roughly 140% over the past 10 years, while the rate of NAS as detected by claims data has increased by 270%.
Anecdotal evidence suggests the true scope of the NAS crisis could be underdetected using traditional diagnostic code surveillance of hospital claims data
A mixed methods evaluation of NAS in Missouri using alternative identification techniques linked to new and expectant mothers in hospital claims data and surveygenerated data support the possibility of significant under-coding of NAS ICD-10 codes P961 and P962
The range of NAS births in Missouri during 2017 varied widely by detection method:
Competing Estimates of the Incidence of NAS Among Missouri Newborns Number of Rate per Surveillance Method Cases 1,000 Births Claims-based by identifying neonates with ICD-10 484 6.7 codes P961 or P962 on discharge record (2017) Claims-based by linking new and expectant mothers to hospital utilization for opioid misuse (2016-2017) Survey-reported and claims-based for non-responding hospitals (2017)
Percent Difference -
3,714
25.4
272%
2,624
35.0
413%
17
