National Forensic Laboratory Information System 2015 Annual Report
National Forensic Laboratory Information System 2015 Annual Report
National Forensic Laboratory Information System 2015 Annual Report
Section 1
National and Regional Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1 .1 Drug Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1 .2 Drug Cases Analyzed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1 .3 National and Regional Drug Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Section 2
Major Drug Categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2 .1 Narcotic Analgesics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2 .2 Tranquilizers and Depressants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2 .3 Anabolic Steroids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2 .4 Phenethylamines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2 .5 Synthetic Cannabinoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Section 3
GIS Analysis: Ethylone and AB-CHMINACA Comparisons,
by Location, 2014 and 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Section 4
Drugs Identified by Laboratories in Selected U .S . Cities . . . . . . . . . . . . . . . . . 22
* Curved trends are sometimes described as U-shaped (i.e., decreasing in earlier years and increasing in recent
years) and S-shaped (i.e., two turns in the trend, roughly either increasing-decreasing-increasing or decreasing-
increasing-decreasing). See Appendix A for a more detailed methodology discussion.
analysis cases. The NFLIS participation rate, defined as the Ventura County Lake Havasu City Santa Fe
Los Angeles County San Bernardino
percentage of the national drug caseload represented by Los Angeles Riverside AZ
Orange County Albuquerque
Long Beach
laboratories that have joined NFLIS, is currently over 97%. Phoenix Scottsdale
San Diego County Phoenix NM
NFLIS includes 50 State systems and 101 local or municipal San Diego PD Mesa
McAllen
San Juan
Mayagez
PR
Ponce
Honolulu
AK
HI
Anchorage
alpha-PVP=alpha-Pyrrolidinopentiophenone
AB-CHMINACA=(N-(1-Amino-3-methyl-1oxobutan-2-yl)-1-(cyclohexylmethyl)1H-indazole-3-carboxamide)
XLR11=[1-(5-Fluoro-pentyl)-1H-indol-3-yl],(2,2,3,3-tetramethylcyclopropyl)methanone
MDMA=3,4-Methylenedioxymethamphetamine
* The estimate for this drug does not meet the standards of precision and reliability. See Appendix A for a more detailed methodology discussion.
1 Sample ns and 95% confidence intervals for all estimates are available on request.
2 As reported by NFLIS laboratories, with no specific drug name provided.
3 Numbers and percentages may not sum to totals because of rounding.
* The estimate for this drug does not meet the standards of precision and
reliability. See Appendix A for a more detailed methodology discussion.
1 Numbers and percentages may not sum to totals because of rounding.
2 Multiple drugs can be reported within a single case, so the cumulative
percentage exceeds 100%. The estimated national total of distinct case
percentages is based on 920,225 distinct cases submitted to State and local
laboratories from January 1, 2015, through December 31, 2015, and
analyzed by March 31, 2016.
60,000 Oxycodone
Hydrocodone
XLR11=[1-(5-Fluoro-pentyl)-1H-indol-3-yl],(2,2,3,3- 50,000
tetramethylcyclopropyl)methanone
40,000
AB-CHMINACA=(N-(1-Amino-3-methyl-1oxobutan-2-yl)-1-
(cyclohexylmethyl)1H-indazole-3-carboxamide) 30,000
20,000
1 Federal drug reports in this table include 28,225 reports from DEA
laboratories and 2,163 reports from CBP laboratories. 10,000
2 Numbers and percentages may not sum to totals because of rounding.
0
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15
Fentanyl
16,000 Clonazepam followed by a decrease through 2015. The most dramatic
decrease in MDMA reports occurred from 2010 to 2012.
12,000
More recently, from 2014 to 2015, reports of cannabis/THC
8,000 (from 437,117 to 395,767 reports) decreased significantly, while
reports of methamphetamine (from 236,175 to 272,823 reports)
4,000 and heroin (from 163,600 to 187,868 reports) increased
significantly (p < .05). The increases in cocaine (from 213,167 to
0
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 216,129 reports) and MDMA (from 4,902 to 5,188 reports)
were not statistically significant.
1 A dashed trend line indicates that estimates did not meet the criteria for
precision or reliability. See Appendix A for a more detailed methodology
discussion.
Figure 1.3 National trend estimates for cannabis/THC and
The S-shaped trend for fentanyl showed that reports methamphetamine, January 2001December 2015
remained steady from 2001 to 2005, which was followed by
700,000
a noticeable increase in 2006. Fentanyl reports continued to
Number of Drug Reports
Heroin
reports) was not statistically significant. 600,000
MDMA
500,000
Other national drug trends
400,000
Figures 1.3 and 1.4 present national trends for reports of
300,000
cannabis/THC, methamphetamine, cocaine, heroin, and
200,000
MDMA. Significant (p < .05) results include the following:
100,000
Cannabis/THC, methamphetamine, and cocaine reports
all showed S-shaped trends. Cannabis/THC decreased 0
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15
from 2001 through 2004, slightly increased from 2005 to
2009, and decreased since 2009. Methamphetamine reports
increased from 2001 through 2005, decreased from 2005
through 2010, then continued to increase from 2011 to 2015.
Cocaine gradually increased from 2001 to 2006, then steadily
decreased through 2014 until a slight increase occurred in
2015.
10 | nflis 2015 annual report
Regional prescription drug trends Figure 1.5 Regional trends in alprazolam reported per
Figures 1.5 through 1.10 show regional trends per 100,000 100,000 persons aged 15 or older, January 2001
persons aged 15 or older for reports of alprazolam, oxycodone, December 20151
hydrocodone, buprenorphine, fentanyl, and clonazepam from 35
2001 to 2015. These figures illustrate changes in prescription West
(per 100,000)
(p< .05) trend results include the following: 20
Note: U.S. census 2015 population data by age were not available for this
publication. Population data for 2015 were imputed.
1 A dashed trend line indicates that estimates did not meet the criteria for
precision or reliability. See Appendix A for a more detailed methodology
discussion.
West
Midwest
Significant (p < .05) trends include the following:
12 Northeast
South For cannabis/THC, the Midwest and South regions showed
(per 100,000)
9
region began around 2008, while the decline in the West
region began around 2009. The Northeast region showed an
6
S-shaped trend, with the trend decreasing from 2011 through
3 2015. The South region showed a linear-decreasing trend.
Between 2014 and 2015, cannabis/THC decreased
0
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 significantly in all regions except the Northeast region (p < .05).
Methamphetamine increased significantly in all regions. Cocaine
increased significantly in the Northeast and Midwest regions.
Figure 1.10 Regional trends in clonazepam reported per Heroin increased significantly in all regions except the South
100,000 persons aged 15 or older, January 2001 region. MDMA increased significantly in the Midwest region
December 2015 only.
15 Heroin fentanyl pills
West
Number of Clonazepam Reports
Midwest
12 Northeast
South
(per 100,000)
0
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15
Note: U.S. census 2015 population data by age were not available for this
publication. Population data for 2015 were imputed.
1 A dashed trend line indicates that estimates did not meet the criteria for
precision and reliability. See Appendix A for a more detailed methodology
discussion.
500 150
Number of Cannabis/THC Reports
West West
Midwest Midwest
(per 100,000)
300 90
200 60
100 30
0 0
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15
Figure 1.12 Regional trends in methamphetamine reported Figure 1.15 Regional trends in MDMA reported per 100,000
per 100,000 persons aged 15 or older, January persons aged 15 or older, January 2001December
2001December 2015 1 2015
Number of Methamphetamine Reports
500 20
West West
Midwest Midwest
Number of MDMA Reports
(per 100,000)
300
10
200
5
100
0 0
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15
500
West
Midwest
Number of Cocaine Reports
400 Northeast
South
(per 100,000)
300
200
100
0
Crack cocaine
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15
Note: U.S. census 2015 population data by age were not available for this
publication. Population data for 2015 were imputed.
1 A dashed trend line indicates that estimates did not meet the criteria for
precision or reliability. See Appendix A for a more detailed methodology
discussion.
40%
19,987
8,144
15,453
3,458
5,896
6,684
6,680
4,690
4,864
3,013
278
a prescription. They are used to treat sleep problems, anxiety, 100% Alprazolam
muscle spasms, and seizures. Even at the recommended doses for Clonazepam
medical treatment, long-term use can lead to physical Diazepam
dependence.iii In 2014, 1.9 million Americans aged 12 or older Phencyclidine (PCP)
used tranquilizers for nonmedical reasons (i.e., without a 80% Other
prescription or simply for the experience or feeling the
tranquilizers caused).iv 26,109
8,901
1,781
6,205
6,107
20%
1,187
2,644
1,617
532
359
Table 2.3 ANABOLIC STEROIDS Figure 2.3 Distribution of anabolic steroid reports within
Number and percentage of anabolic steroid reports region, 20151
in the United States, 20151 West Midwest Northeast South
Anabolic Steroid Reports Number Percent Testosterone
100%
Testosterone 1,905 48.43% Trenbolone
Number and Percentage of Anabolic Steroid Reports
Drostanolone 70 1.77%
Mestanolone 34 0.86%
286
275
Mesterolone 26 0.66%
40%
249
Methenolone 19 0.47%
Dehydroepiandrosterone 16 0.40%
257
174
375
4-Chlorodehydromethyltestosterone 11 0.29%
129
82
136
89
63
53
54
102
43
111,124
phenethylamines reported in the Northeast region were
47,639
109,507
Number and Percentage of Phenethylamine Reports
Ethylone
amphetamine. The Northeast region also reported the highest alpha-PVP
percentages of ethylone (16%) and alpha-PVP (7%). 80% Other
Table 2.4 PHENETHYLAMINES
Number and percentage of phenethylamine reports in
the United States, 20151
60%
Phenethylamine Reports Number Percent
4,554
Methamphetamine 272,823 86.21%
Amphetamine 12,222 3.86%
Ethylone 9,237 2.92% 40%
alpha-PVP * *
MDMA 5,188 1.64%
Lisdexamfetamine 1,941 0.61%
1,547
1,486
1,451
665
6,340
6,173
5,028
3,071
955
898
562
233
Other phenethylamines 2,178 0.69% December 31, 2015, that were analyzed by March 31, 2016.
2 Numbers and percentages may not sum to totals because of rounding.
8,693
A total of 33,820 synthetic cannabinoid reports were
2,527
identified during 2015, accounting for about 2% of all drugs
2,170
reported (Table 2.5). AB-CHMINACA (22%) and XLR11 40%
993
(21%) were the most commonly identified synthetic
855
1,520
4,482
769
cannabinoids, followed by AB-PINACA (7%),
1,361
1,207
3,323
AB-FUBINACA (7%), and 5-fluoro AMB (7%).
1,027
20%
AB-CHMINACA accounted for one-fifth or more of all
798
647
324
559
490
1,143
synthetic cannabinoid reports in the West region (29%), South
881
region (24%), and Northeast region (20%) (Figure 2.5). XLR11
52
accounted for one-fifth or more of all synthetic cannabinoids 0% Total Number2
reported in the West region (26%), Northeast region (25%), and 2,994 6,271 6,034 18,521 33,820
Midwest region (22%). The Midwest region reported the
highest percentage of AB-PINACA (13%). In the Northeast AB-CHMINACA=(N-(1-Amino-3-methyl-1oxobutan-2-yl)-1-
(cyclohexylmethyl)1H-indazole-3-carboxamide)
region, 11% of synthetic cannabinoids were reported as
XLR11=[1-(5-Fluoro-pentyl)1H-indol-3-yl],(2,2,3,3-
AB-FUBINACA. tetramethylcyclopropyl)methanone
AB-PINACA=(N-(1-Amino-3-methyl1-oxobutan-2-yl)-1-pentyl-1H-
indazole3-carboxamide)
Table 2.5 SYNTHETIC CANNABINOIDS AB-FUBINACA=(N-(1-Amino-3-methyl-1-oxobutan-2-yl)-1-(4-
fluorobenzyl)-1H-indazole-3-carboxamide)
Number and percentage of synthetic cannabinoid
5-fluoro AMB=methylN-{[1-(5-fluoropentyl)-1H-indazol-3-yl]carbonyl}
reports in the United States, 20151 valinate
Synthetic Cannabinoid Reports Number Percent MAB-CHMINACA=N-(1-Amino-3,3-dimethyl-1-oxobutan-2-yl)-1-
(cyclohexylmethyl)-1H-indazole-3-carboxamide
AB-CHMINACA 7,571 22.39% NM2201=Naphthalene-1-yl 1-(5-fluoropentyl)-1H-indole-3-carboxylate
XLR11 6,973 20.62% 5-fluoro ADB=Methyl (R)-2-(1-(5-Fluoropentyl)-1H-indazole-3-
AB-PINACA 2,493 7.37% carboxamido)-3,3-dimethylbutanoate
AB-FUBINACA 2,402 7.10% 5F-AB-PINACA=N-(1-Amino-3-methyl-1-oxobutan-2-yl)-1-(5-
fluoropentyl)-1H-indazole-3-carboxamide
5-fluoro AMB 2,258 6.68% FUB-AMB=Methyl 2-({1-[(4-fluorophenyl)methyl]-1H-indazole-3-
MAB-CHMINACA 1,711 5.06% carbonyl}amino)-3-methylbutanoate
NM2201 1,332 3.94% ADB-FUBINACA=N-(1-Amino-3,3-dimethyl-1-oxobutan-2-yl)-1-(4-
5-fluoro-ADB 805 2.38% fluorobenzyl)-1H-indazole-3-carboxamide
5F-AB-PINACA 538 1.59% 5F-PB-22=(Quinolin-8-yl 1-(5-fluoropentyl)-1H-indole-3-carboxylate)
FUB-AMB 534 1.58% FUB-PB-22=Quinolin-8-yl 1-(4-fluorobenzyl)-1H-indole-3-carboxylate
ADB-FUBINACA 450 1.33% MDMB-FUBINACA=Methyl (S)-2-(1-(4-fluorobenzyl)-1H-indazole-
3-carboxamido)-3,3-dimethylbutanoate
5F-PB-22 408 1.21% AKB48 N-(5-fluoropentyl)=N-(1-adamantyl)-1-(5-fluoropentyl)-1H-
FUB-PB-22 343 1.01% indazole-3-carboxamide
MDMB-FUBINACA 328 0.97%
vii Office of National Drug Control Policy, The White House. (n.d.).
AKB48 N-(5-fluoropentyl) 313 0.92%
Other synthetic cannabinoids 5,361 15.85% Synthetic drugs (a.k.a. K2, spice, bath salts, etc.). Retrieved from http://
www.whitehouse.gov/ondcp/ondcp-fact-sheets/synthetic-drugs-k2-
Total Synthetic Cannabinoid Reports2 33,820 100.00% spice-bath-salts
viii Law, R., Schier, J., Martin, C., Chang, A., & Wolkin, A. (2015, June
Total Drug Reports 1,549,466
12). Notes from the field: Increase in reported adverse health effects
1 Includes drugs submitted to laboratories from January 1, 2015, through related to synthetic cannabinoid use United States, JanuaryMay
December 31, 2015, that were analyzed by March 31, 2016. 2015. Morbidity and Mortality Weekly Report, 64, 618619. Retrieved
2 Numbers and percentages may not sum to totals because of rounding.
from https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6422a5.
htm
18 | nflis 2015 annual report
Section 3 GIS ANALYSIS:
ETHYLONE AND
AB-CHMINACA
COMPARISONS,
BY LOCATION,
2014 AND 2015
One of the unique features of This section presents data at the State and county levels for
NFLIS is the ability to analyze and the percentage of drug reports identified as ethylone and
AB-CHMINACA at two points in time2014 and 2015.
monitor, by the county of origin,
Reports of ethylone and AB-CHMINACA increased
variation in drugs reported by substantially in NFLIS between 2014 and 2015. Ethylone was
laboratories. By using Geographic first reported NFLIS in 2011, and AB-CHMINACA was first
Information System (GIS) analyses, reported in 2014. In 2015, both drugs first appeared in the
NFLIS can provide information on NFLIS list of the top 25 most frequently identified drugs;
ethylone and AB-CHMINACA were the 12th and 14th most
drug seizure locations.
frequently reported drugs, respectively.
The GIS data presented here are based on information
provided to NFLIS forensic laboratories by the submitting law
enforcement agencies (Figures 3.1 to 3.8). The information
submitted by law enforcement includes the ZIP Code or county
of origin associated with the drug seizure incident or the name
of the submitting law enforcement agency. When a ZIP Code or
county of origin is unavailable, the drug seizure or incident is
assigned to the same county as the submitting law enforcement
agency. If the submitting agency is unknown, the seizure or
incident is assigned to the county in which the laboratory
completing the analyses is located.
It is important to note that these data may not include all
drug items seized at the State and county levels. Instead, these
data represent only those drugs that were submitted to and
analyzed by NFLIS forensic laboratories. In addition, some
laboratories within several States are not currently reporting data
to NFLIS, and their absence may affect the relative distribution
of drugs seized and analyzed. Nevertheless, these data can serve
as an important source for identifying abuse and trafficking
trends and patterns across and within States.
Figure 3.3 Percentage of total drug reports identified as Figure 3.4 Percentage of total drug reports identified as
AB-CHMINACA, by State, 20141 AB-CHMINACA, by State, 20151
1 Includes drugs submitted to State and local laboratories during the calendar year that were analyzed within three months of the reporting period.
231
231
10 Tallahassee
(
!
(
! Jackson ville
10 Tallahassee
(
!
(
! Jackson ville
9
9
417
417
Orla ndo !
(
528 Orla ndo !
(
528
!( 275
!( 275
Tampa Tampa
175
175
91
91
95
95
27
27
869
869
Miami !
( Miami !
(
5 .013 .6 5 .07 .9
3 .04 .9
821
3 .04 .9
821
1 .02 .9 1 .02 .9
0 .10 .9 0 .10 .9
0 .0 0 .0
No Data No Data
Figure 3.7 Percentage of total drug reports identified as Figure 3.8 Percentage of total drug reports identified as
AB-CHMINACA in Pennsylvania, by county, AB-CHMINACA in Pennsylvania, by county,
20141 20151
(
! (
!
Erie
90
Erie
90
79
79
84
84
180
380
180
380
80
80
Allentown !
!(
279
(
!(
279
Allentown !
(
Pittsburgh
476
Pittsburgh
476
99
76
176
99
176
276
70
81
(
!
70
81
76 276
(
!
Philadelphia
83
95
83
95
Philadelphia
1 Includes drugs submitted to State and local laboratories during the calendar year that were analyzed within three months of the reporting period.
50% 100%
100%
100% 0% 50%
Seattle
50%
50% 0% 100%
Spokane
0%
0% 50% Minneapolis-St. Paul
Portland
0%
100% 100% Rapid City 100%
100%
50% 50% 50%
100%
100% 50%
0% 0% 0%
Sacramento Cheyenne 50% Chicago
100% 50% 0%
Des Moines
100% 0%
50% 0% 100%
Lincoln
100% Salt Lake City
50% 100%
0% 50% 100%
San Francisco 50%
0% 50%
Las Vegas 0% 50%
100% 0% Denver
Fresno 0%
0% St. Louis
50% 100% Topeka
100% 100%
1
0% 50% 50%
Los Angeles 50% 100%
0% 0% 0%
San Diego Phoenix 50%
Sante Fe 100%
0%
50% Dallas
Note: Based on the total number of drugs reported,
0%
drugs that were reported less than 2% are not El Paso
presented even if they were one of the top four Cannabis/THC Hydrocodone
Cocaine Oxycodone
drugs for a selected location. Data reported for some
Methamphetamine AB-CHMINACA 100%
laboratories, especially State system laboratories, may XLR11
Heroin
include data from areas outside the referenced city. Fentanyl
Alprazolam 50%
0%
McAllen
22 | nflis 2015 annual report
The highest percentages of methamphetamine were reported by Selected Laboratories
laboratories representing cities in the West and Midwest, such as Fresno Atlanta (Georgia State Bureau of InvestigationDecatur Laboratory)
(64%), Rapid City (52%), Portland (51%), Sacramento (50%), San Diego Augusta (Maine Department of Human Services)
(50%), Lincoln (49%), Spokane (44%), Minneapolis-St. Paul (44%), Baltimore (Baltimore City Police Department)
Baton Rouge (Louisiana State Police)
LosAngeles (37%), Des Moines (31%), and Santa Fe (31%). Cities in the
Birmingham (Alabama Department of Forensic SciencesBirmingham
South, such as Dallas (41%), Oklahoma City (38%), and Atlanta (30%), also Laboratory)
reported a high percentage of drugs identified as methamphetamine. Cheyenne (Wyoming State Crime Laboratory)
Nationally, 18% of drugs in NFLIS were identified as methamphetamine. Chicago (Illinois State PoliceChicago Laboratory)
The highest percentages of heroin were reported by laboratories Cincinnati (Hamilton County Coroners Office)
representing the Northeastern cities of Pittsburgh (39%) and Augusta (33%); Columbia (South Carolina Law Enforcement DivisionColumbia
Laboratory)
the Midwestern cities of Cincinnati (26%), St. Louis (22%), and Chicago
Dallas (Texas Department of Public SafetyGarland Laboratory)
(22%); the Southern cities of Baltimore (27%) and Louisville (23%); and the
Denver (Denver Police Department Crime Laboratory)
Western cities of Seattle (25%), Phoenix (24%), and Portland (22%).
Des Moines (Iowa Division of Criminal Investigations)
Nationally, 12% of all drugs in NFLIS were identified as heroin.
El Paso (Texas Department of Public SafetyEl Paso Laboratory)
Among controlled prescription drugs, Nashville (6%) reported the highest Fresno (California Department of JusticeFresno Laboratory and Fresno
percentage of oxycodone, followed by Las Vegas (5%) and Philadelphia (5%). County Sheriffs Forensic Laboratory)
Houston (Texas Department of Public SafetyHouston Laboratory and
Nationally, 3% of drugs in NFLIS were identified as oxycodone. Little Rock Harris County Medical Examiners Office)
(5%) and Jackson (5%) reported the highest percentages of hydrocodone. Indianapolis (Indianapolis-Marion County Forensic Laboratory)
Nationally, 2% of drugs in NFLIS were identified as hydrocodone. McAllen Jackson (Mississippi Department of Public SafetyJackson Laboratory
(8%) and Tampa (6%) reported the highest percentages of alprazolam. and Jackson Police Department Crime Laboratory)
Nationally, 3% of drugs in NFLIS were identified as alprazolam. Cincinnati Las Vegas (Las Vegas Metropolitan Police Crime Laboratory)
(9%) and Augusta (6%) reported the highest percentages of fentanyl, while Lincoln (Nebraska State Patrol Criminalistics LaboratoryLincoln
Laboratory)
Salt Lake City (6%) reported the highest percentage of XLR11. Less than 1%
Little Rock (Arkansas State Crime Laboratory)
of drugs in NFLIS were identified as fentanyl or XLR11.
Los Angeles (Los Angeles Police Department and Los Angeles County
Sheriffs Department)
Louisville (Kentucky State PoliceLouisville Laboratory)
McAllen (Texas Department of Public SafetyMcAllen Laboratory)
100%
Miami (Miami-Dade Police Department Crime Laboratory)
50% Minneapolis-St. Paul (Minnesota Bureau of Criminal Apprehension
Minneapolis Laboratory)
0%
Augusta Montgomery (Alabama Department of Forensic SciencesMontgomery
Laboratory)
Nashville (Tennessee Bureau of InvestigationNashville Laboratory)
100% 100%
New York City (New York City Police Department Crime Laboratory)
100% 100%
50% 50% Oklahoma City (Oklahoma State Bureau of InvestigationOklahoma City
50% Laboratory)
50%
0% 0%
Pittsburgh Orlando (Florida Department of Law EnforcementOrlando Laboratory)
New York City
0% 0%
100% Cincinnati Philadelphia Philadelphia (Philadelphia Police Department Forensic Science
100% Laboratory)
100%
50% Phoenix (Phoenix Police Department)
50%
100% 50% Pittsburgh (Allegheny County Coroners Office)
0% 100%
Indianapolis 0% Portland (Oregon State Police Forensic Services DivisionPortland
Louisville 50% 0%
100% 50% Baltimore Laboratory)
0%
Rapid City (Rapid City Police Department)
50% 0% 100% Raleigh Raleigh (North Carolina State Bureau of InvestigationRaleigh
100% Nashville
100% 100% Laboratory)
0% 50%
50% Little Rock
50% Sacramento (Sacramento County District Attorneys Office)
50%
100% 0% Salt Lake City (Utah State Crime LaboratorySalt Lake City Laboratory)
0% Atlanta
Oklahoma City 0% 0%
Columbia San Diego (San Diego Police Department)
50% Birmingham 100%
San Francisco (San Francisco Police Department)
100% 0% 50% Santa Fe (New Mexico Department of Public SafetySanta Fe Laboratory)
100% Jackson
50% 100% 0% Seattle (Washington State PatrolSeattle Laboratory)
50% Orlando 100% Spokane (Washington State PatrolSpokane Laboratory)
0% 50% 100%
Baton Rouge St. Louis (St. Louis Police Department)
0% 50%
Houston 0% 50% Tampa (Florida Department of Law EnforcementTampa Laboratory)
Montgomery
0% Topeka (Kansas Bureau of InvestigationTopeka Laboratory)
0% Miami
Tampa
nflis 2015 annual report | 23
Appendix A STATISTICAL METHODOLOGY
where
= set of all nonmissing months in laboratory ,
= case count for laboratory in month , and
= mean case counts for all laboratories reporting
ix complete data.
The case and item loads for the nonsampled laboratories were used
in calculating the weights.
x In the current reporting period, for example, out of 112 nonsampled
laboratories and laboratory systems, 80 (or 72%) reported.
where Each weight has two components, the design weight and the
nonresponse adjustment factor, the product of which is the final
= set of all nonmissing months in laboratory , weight used in estimation. After imputation, the final item weight
= item count for laboratory in month , and is based on the item count, and the final case weight is based on
= case count for laboratory in month . the case count of each laboratory or laboratory system. The final
weights are used to calculate national and regional estimates. The
Drug-specific case and report counts are imputed using the
first component, the design weight, is based on the proportion of
same imputation techniques presented above for the case and
the caseload and item load of the NFLIS universexi represented
item counts. The total drug, item, and case counts are calculated
by the individual laboratory or laboratory system. This step takes
by aggregating the laboratory and laboratory system counts for
advantage of the original PPS sample design and provides precise
those with complete reporting and those that require imputation.
estimates as long as the drug-specific case and report counts are
correlated with the overall caseload and item load.xii
NEAR imputations and drug report-level
adjustments For noncertainty reporting laboratories in the sample (and
reporting laboratories in the certainty strata with nonreporting
Most forensic laboratories classify and report case-level
laboratories), the design-based weight for each laboratory is
analyses consistently in terms of the number of vials of a
calculated as follows:
particular pill. A small number, however, do not produce drug
report-level counts in the same way as those submitted by the vast
majority. Instead, they report as items the count of the individual
pills themselves. Laboratories that consider items in this manner
also consider drug report-level counts in this same manner. Drug where
report-to-case ratios for each drug were produced for the similarly = th laboratory or laboratory system;
sized laboratories, and these drug-specific ratios were then used to = sum of the case (item) counts for all of the
adjust the drug report counts for the relevant laboratories. laboratories and laboratory systems (sampled and
nonsampled) within a specific stratum, excluding
NEAR weighting procedures certainty strata and the volunteer stratum; and
Each NFLIS reporting laboratory was assigned a weight = number of sampled laboratories and laboratory
systems within the same stratum, excluding
to be used in calculating design-consistent, nonresponse-adjusted
certainty strata and the volunteer stratum.
estimates. Two weights were created: one for estimating cases
and one for estimating drug reports. The weight used for case Certainty laboratories were assigned a design weight of one.xiii
estimation was based on the caseload for every laboratory in
the NFLIS population, and the weight used for drug reports
estimation was based on the item load for every laboratory in xi
See the Introduction of this publication for a description of the
the NFLIS population. For reporting laboratories, the caseload NFLIS universe.
xii
and item load used in weighting were the reported totals. For Lohr, S. L. (2010). Sampling: Design and analysis (2nd ed., pp. 231-
nonreporting laboratories, the caseload and item load used in 234). Boston, MA: Brooks/Cole.
xiii
weighting were obtained from an updated laboratory survey With respect to the design weight, reporting laboratories and
administered in 2013. laboratory systems in certainty strata with nonreporting laboratories
and laboratory systems are treated the same way as reporting
When the NFLIS sample was originally drawn, two stratifying noncertainty sampled laboratories and laboratory systems. This is
done to reduce the variance; otherwise, all reporting laboratories
variables were used: (1) type of laboratory (State system or and laboratory systems in these strata would get the same weight
municipal or county laboratory) and (2)determination of regardless of their size.
For some drugs, such as cannabis/THC and cocaine, var( T2014 ) = variance of T2014 ; and
thousands of reports occur annually, allowing for reliable national
prevalence estimates to be computed. For other drugs, reliable and cov( T2014 ,T2015 ) = covariance between T2014 and T2015 .
precise estimates cannot be computed because of a combination
of low report counts and substantial variability in report counts For the national prior-year comparisons, a = b = 1. For the
between laboratories. Thus, a suppression rule was established. regional prior-year comparisons, a = 100,000 divided by the
Precision and reliability of estimates are evaluated using the regional population total for 2015, and b = 100,000 divided by the
relative standard error (RSE), which is the ratio between the regional population total for 2014.
standard error of an estimate and the estimate. Drug estimates The percentile of the test statistic in the t distribution
with an RSE > 50% are suppressed and not shown in the tables. determined whether the prior-year comparison was statistically
significant (a two-tailed test at = .05).
xiv See footnote xiii.
where ,
is a 15 1 vector of errors due to the probability
sample, and where
=15 1 vector of errors due to the underlying model. for all , and
provide contributions for the first-
Randomness due to the sample exists because only a sample of
order (linear), second-order (quadratic), and third-order
all eligible laboratories has been randomly selected to be included.
(cubic) polynomials, respectively.
Randomness due to the population exists because many factors
that can be viewed as random contribute to the specific total Note that the error term is the same in the original model
reported by a laboratory in a time period. For example, not all and the reparameterized model because the fitted surface is the
drug seizures that could have been made were actually made, and same for both models. The model was further constrained to
there may have been some reporting errors. If rates (per 100,000 have regression residuals sum to zero, a constraint that is not
persons aged 15 years or older) and not totals are of interest, the guaranteed by theory for these models but was considered to
above model can be applied to , where c equals 100,000 improve model fit due to an approximation required to estimate
divided by the 15-or-older regional population size as given by S . Standard errors of the regression trend estimates were
the U.S. Census Bureau. obtained by simulation.
The regression model used to perform the analysis is Final models were selected after testing for the significance of
coefficients at the = 0.05 level (p < .05), which means that if the
, trend of interest (linear, quadratic, cubic) was in fact zero, then
there would be a 5% chance that the trend would be detected as
where statistically significant when in fact it is not. Final fitted models
Yt = the population total value, considered to be a are most easily interpreted using graphical plots.
realization of the underlying model; and
t = one of a set of 15 independent normal variates with a
mean of zero and a variance of .
Benefits Limitations
The systematic collection and analysis of drug analysis data NFLIS has limitations that must be considered when
aid our understanding of the Nations illicit drug problem. interpreting findings generated from the database.
NFLIS serves as a resource for supporting drug scheduling
policy and drug enforcement initiatives nationally and in specific Currently, NFLIS includes data from Federal, State, and local
communities around the country. forensic laboratories. Federal data are shown separately in this
publication. Efforts are under way to enroll additional Federal
Specifically, NFLIS helps the drug control community achieve laboratories.
its mission by
NFLIS includes drug chemistry results from completed
providing detailed information on the prevalence and types of analyses only. Drug evidence secured by law enforcement but
controlled substances secured in law enforcement operations; not analyzed by laboratories is not included in the database.
identifying variations in controlled and noncontrolled National and regional estimates may be subject to variation
substances at the national, State, and local levels; associated with sample estimates, including nonresponse bias.
identifying emerging drug problems and changes in drug State and local policies related to the enforcement and
availability in a timely fashion; prosecution of specific drugs may affect drug evidence
monitoring the diversion of legitimately marketed drugs into submissions to laboratories for analysis.
illicit channels; Laboratory policies and procedures for handling drug evidence
providing information on the characteristics of drugs, including vary. Some laboratories analyze all evidence submitted to
quantity, purity, and drug combinations; and them, whereas others analyze only selected case items. Many
laboratories do not analyze drug evidence if the criminal case
supplementing information from other drug sources, including was dismissed from court or if no defendant could be linked to
the National Survey on Drug Use and Health (NSDUH) and the case.
the Monitoring the Future (MTF) study.
Laboratories vary with respect to the records they maintain.
NFLIS is an opportunity for State and local laboratories to For example, some laboratories automated records include the
participate in a useful, high-visibility initiative. Participating weight of the sample selected for analysis (e.g., the weight of
laboratories regularly receive reports that summarize national
one of five bags of powder), whereas others record total weight.
and regional data. In addition, the Data Query System (DQS)
is a secure website that allows NFLIS participantsincluding
State and local laboratories, the DEA, and other Federal drug
control agenciesto run customized queries on the NFLIS data.
Enhancements to the DQS provide a new interagency exchange
forum that will allow the DEA, forensic laboratories, and other
members of the drug control community to post and respond to
current information.
The NFLIS website (https://www.nflis.deadiversion.usdoj. The private site requires user accounts, and security roles
gov/) is an important feature of the NFLIS program. It is the are assigned to manage access to its features, including the
key resource to provide NFLIS-related information, through a Map Library, NFLIS Data Entry Application, and DQS. The
public site and through a private site, which gives secure access to DQS is a distinct resource for NFLIS reporting laboratories to
the NFLIS DQS. run customizable queries on their own case-level data and on
The public site is frequently updated with NFLIS-related aggregated metropolitan, State, regional, and national data.
news, including information relevant to drug control efforts Features include the drug category queries for synthetic
and DEA participation in conferences. Also available are cannabinoids and synthetic cathinones.
downloadable versions of published NFLIS reports, links to
other websites, and contact information for key NFLIS staff.
Public features include a link to the Scientific Working Group To obtain information about NFLIS participation
for the Analysis of Seized Drugs (SWGDRUG) mass spectral or the DQS, please visit the NFLIS website at
library at http://www.swgdrug.org/. https://www.nflis.deadiversion.usdoj.gov/.
September 2016