Data on the CNICS cohort are harmonized in multiple domains across ten contributing sites as detailed below. Data across all domains are updated every four months in the CNICS Research Platform.

Demographic

  • CNICS captures demographic data including sex, birth year, self-identified race, self-identified Hispanic origin, transgender status, and country of birth (see table below)
  • Sex, age, race and Hispanic ethnicity data are collected using Health Resources and Services Administration standard coding (age distribution of the CNICS cohort 1995-2020 shown in figure below)
  • Risk factors for HIV acquisition are coded according to the 1993 Centers for Disease Control and Prevention classification system (1993 Revised Classification System for HIV Infection and Expanded Surveillance Case Definition for AIDS Among Adolescents and Adults. MMWR Recomm Rep 1992;41(RR-17):1-19)

Demographic characteristics of the CNICS Cohort

Age at enrollment

<20

301

1%

20-29

6936

19%

30-39

12350

33%

40-49

11199

30%

50-59

5131

14%

>59

1272

3%

Sex

Male

30478

82%

Female

6711

18%

Transgender

466

1%

Race

White

18408

50%

Black

14521

39%

Asian/ Pacific Islander

762

2%

Multiracial

1052

3%

Other

1569

4%

Unknown

877

2%

Hispanic ethnicity

No

29108

78%

Yes

5176

14%

Unknown

2905

8%

Risk factors for HIV acquisition

MSM*

19965

54%

Injection drug users (IDU)

3710

10%

MSM and IDU

2487

7%

Heterosexual contact

8929

24%

Other

1077

3%

Unknown

1021

3%

*Men who have sex with men 

 

Age Distribution in the CNICS Cohort by Year (1995-2020)
57% are 50 years or older in 2020

Diagnosis Data

  • CNICS captures diagnoses prospectively recorded in the Electronic Health Record by treating clinicians during routine care; historical diagnoses are captured at a patient’s initial visit to a CNICS site
  • Diagnosis data are mapped to the CNICS standard diagnosis codes and harmonized across CNICS sites
  • Diagnoses in CNICS are 1. verified through protocol-driven medical record review and centrally adjudicated by expert panels of physicians (e.g. stroke ascertainment and adjudication, characteristics of PLWH with Type 1 vs Type 2 myocardial infarction) or 2. confirmed by CNICS Operational Definitions that include laboratory test results and medication data (e.g. diabetes identification in EHR data)

1. Verified diagnoses

*Medical record review; **Adjudication

AIDS-Defining Illnesses (ADI)*

ADIs are verified in accordance with the “CDC 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults”(1)

End stage liver disease (ESLD)*

ESLD diagnoses verified through 2015 (ascites, spontaneous bacterial peritonitis, esophageal/gastric variceal hemorrhage, hepatic encephalopathy, hepatocellular carcinoma)

Cancer*

Cancer diagnoses (AIDS-defining and non-AIDS defining) verified through 2016

Myocardial infarction (MI)**

Over 1,300 adjudicated MIs (Type 1 or Type 2) to date

Congestive heart failure (CHF)

CHF adjudication ongoing

Stroke

Over 450 adjudicated strokes to date

Venous thromboembolism (VTE)**

Over 450 adjudicated VTEs to date

2. Confirmed diagnoses

Anemia

Hematocrit/hemoglobin values

Chlamydia trachomatis

Laboratory-based testing

Chronic Kidney disease

Creatinine values, eGFR (CKD-EPI, other equations)

Chronic Liver disease

Laboratory data including FIB-4

COPD

Pulmonary medication/inhaler data

Diabetes mellitus

Diabetes-related medication, laboratory, and diagnosis data

Dyslipidemia

Lipid-related medication and laboratory data

Neisseria gonorrhea

Laboratory-based testing

Hepatitis B virus

HBV serology/DNA data

Hepatitis C virus

HCV serology/RNA data

Hypertension

Hypertension-related medication, blood pressure, and diagnosis data (treated vs untreated, controlled vs. uncontrolled)

Syphilis

Syphilis serology data

See examples of CNICS stroke and MI data below

1. Adjudicated Strokes and Ischemic Stroke Subtypes in the CNICS Cohort

2. Adjudicated MIs in the CNICS Cohort by Type (50% Type 1 MI, 50% Type 2 MI)
(adapted from Thygesen K, et al. J Am Coll Cardiol. 2012)

3. Causes of Type 2 MIs in the CNICS Cohort
(adapted from Crane H, et al. JAMA Cardiol. 2017) 

Abbreviations: HIV, human immunodeficiency virus; T2MI, type 2 myocardial infarction events
*Noncoronary cardiac causes include nonatherosclerotic causes, such as those associated with congestive heart failure and cardiac tumor
†Hypotension not owing to sepsis, gastrointestinal bleeding, drug overdose, or other listed causes
‡Events that occur in the setting of noncardiac procedures, such as abdominal surgery and lower extremity amputation

Laboratory Data and Vital Signs

  • Laboratory test results are uploaded directly from clinical laboratory medicine systems and harmonized across CNICS sites implementing standard units and clinical interpretations
  • Vital signs captured in CNICS include blood pressure and height and weight measurements that are used to compute BMI (see example in figure below)
  • Examples of CNICS laboratory data are shown in the table below and provided in detail here (click table to enlarge)

CNICS Laboratory Data Examples

C-reactive protein

Cardiac biomarkers (e.g. troponin)

CD4, CD8 absolute/% etc.

Complete Blood Count (CBC)

Chemistries

Chlamydia trachomatis

Coagulation (PT/PTT/INR)

Cryptococcal antigen

Epstein-Barr Virus (EBV)

Hemoglobin A1c

Hepatitis A

Hepatitis B

Hepatitis C

Hepatitis Delta

Herpes simplex virus (HSV)

HIV-1 antibody/antigen

HIV-1 viral load (all assays over time)

HIV-1, HIV-2 antibody

Human Leukocyte Antigen – HLA-B5701

Lipids

Liver function tests

Neisseria gonorrhea

Prostate specific antigen

Quantiferon TB gold

Syphilis

Testosterone

Toxoplasma antibody

Trichomonas

Urinalysis

Vitamin D

Vital signs

Systolic/Diastolic blood pressure

Height

Weight

See examples of CNICS HIV viral suppression, HCV viral load, and BMI data in figures below

1. Viral Suppression by Year in the CNICS Cohort

2. HCV Viral Load Status by Year in the CNICS Cohort

3. Transitions Between BMI Categories Following Initiation of Dolutegravir-based ART Regimens

Sankey diagram of changes in BMI category between antiretroviral therapy (ART) initiation [baseline] and 2 years post ART initiation [follow-up] among previously ART-naïve people living with HIV (PLWH) initiating dolutegravir (DTG)-based integrase strand transfer inhibitor (INSTI) ART regimens. In the first 2 years of treatment, a greater proportion of PLWH shifted to a higher BMI category compared to those who shifted to a lower BMI category.

Medication Data

  • Medications prescribed, including start/stop dates, are entered into the Electronic Health Record by treating clinicians and used to compute courses of therapy
  • Medications including antiretroviral regimens and Direct Acting Antiviral (DAA) drugs are verified through medical record review (see examples of CNICS ART and DAA data in figures below)

CNICS Medication Data Examples

Anabolic steroids

Anti-hypertensives

Corticosteroids

Antibiotics

Antifungals

Antimalarials

Antitubercular

Anticoagulants

Direct-Acting Antiviral Agents

Antiretrovirals (e.g. NRTI, NNRTI, PI, INSTI)

Pulmonary medications (e.g. inhaled, oral)

Diabetes-related (e.g. insulin, oral agents)

Hormones (e.g. testosterone, estrogen)

Lipid lowering

Psychiatric (anti-anxiety, depressant, psychotic,
mood stabilizer)

Opioids

Antiepileptic

Substance use treatment 

 

1. ART Regimens by Drug Class in CNICS Cohort (2018-19)
*ex. Dolutegravir/Darunavir

2. DAA Treatment Response in CNICS Cohort
(adapted from Kim H, et al. OFID. 2019) 

 

Antiretroviral Resistance Data

  • CNICS captures viral resistance data including full nucleotide genotype, phenotype, and tropism assays and has the capability for expansion to include new drug targets
  • CNICS genotypic resistance data are processed using the Stanford HIV Drug Resistance Database
  • Study demonstrating Substantial Decline in Heavily Treated Therapy Experienced Persons with HIV with Limited Antiretroviral Treatment Options (see abstract and figures below; PubMed link here)

Objective: Historically, a high burden of resistance to antiretroviral therapy (ART) in heavily treatment-experienced (HTE) persons with HIV (PWH) resulted in limited treatment options (LTOs). We evaluated the prevalence, risk factors, and virologic control of HTE PWH with LTO throughout the modern ART era.

Design: We examined all ART-experienced PWH in care between 2000 and 2017 in the Centers for AIDS Research Network of Integrated Clinical Systems cohort.

Methods: We computed the annual prevalence of HTE PWH with LTO defined as having two or less available classes with two or less active drugs per class based on genotypic data and cumulative antiretroviral resistance. We used multivariable Cox proportional hazards models to examine risk of LTO by 3-year study entry periods adjusting for demographic and clinical characteristics.

Results: Among 27 133 ART-experienced PWH, 916 were classified as having LTO. The prevalence ofPWH with LTO was 5.2–7.5% in 2000–2006, decreased to 1.8% in 2007, and remained less than 1% after 2012. Persons entering the study in 2009–2011 had an 80%lower risk of LTOcompared with those entering in2006–2008 (adjusted hazard ratio 0.20; 95%confidence interval: 0.09–0.42). We found a significant increase in undetectable HIV viral loads among PWH ever classified as having LTO from less than 30% in 2001 to more than 80% in 2011, comparable with persons who never had LTO.

Conclusion: Results of this large multicenter study show a dramatic decline in the prevalence of PWH with LTO to less than 1% with the availability of more potent drugs and a marked increase in virologic suppression in the current ART era.

1. Annual Prevalence of PWH with Limited Treatment Options (LTO) Among ART-experienced Persons in Care by Year (2000–2017)

2. Percentage of Undetectable HIV Viral Load Tests by Year Among Antiretroviral-experienced PWH by Limited Treatment Option (LTO) Status (2000-2017)

AIDS 2020, 34:2051–2059

Patient Reported Measures and Outcomes (PROs)

  • PROs are collected at CNICS sites using validated survey instruments (see references here) administered at routine clinical care visits with results available for use by clinicians at the time of the encounter 
  • Patients complete PRO assessments every four to six months using touch-screen tablets connected to a wireless network with SSL/TLS encryption
  • Over 80,000 PRO assessments have been completed by over 21,000 patients in CNICS
  • Recent expansion of PRO domains include the collection of e-cigarette/vaping, illicit fentanyl use, access to Narcan among illicit opioid users, and partner use of PrEP
  • New approaches to PRO collection include the ability to complete PRO assessments remotely in response to the shift to Telehealth visits during the COVID-19 pandemic

CNICS PRO Domain

Instrument

Symptoms

Depression and anxiety

Patient Health Questionnaire (PHQ-9, PHQ-5)[1,2]

Symptom burden (20 symptoms)

HIV Symptom Index[3]

Behaviors

Adherence to antiretroviral therapy

Visual analog scale (VAS), self-rating item, and items from AACTG[4-6]

Smoking including e-cigarette/vaping use

Tobacco instrument with expansion to include e-cigarette/vaping

Alcohol use

Alcohol Use Disorders Identification Test (AUDIT-C), full AUDIT annually among those at risk on AUDIT-C, Alcohol Mini International Neuropsychiatric Interview (MINI) among those at-risk on AUDIT[7-9]

Drug use: includes type, mode, frequency, impact, overdoses, Narcan supply, needle sharing, and drug treatment

Modified Alcohol, Smoking and Substance Involvement Screening Test (ASSIST)[10,11] and drug treatment items adapted from Treatment Services Review[12]

Sexual risk behaviors

Sexual Risk Behavior Inventory[13]

Body shape and activity

Physical activity level

Lipid Research Clinical Questionnaire (LRCQ)[3]

Body morphology

Fat Redistribution and Metabolism (FRAM)[14]

Frailty

Composite definition including items from multiple other instruments

Identify

Sexual orientation

 

Sexual identity

 

Violence

Intimate Partner Violence 

Intimate partner violence 4-item measure (IPV-4)[15]

Childhood household violence

Adapted from Adverse Childhood Experiences-International Questionnaire (ACE-IQ)[16]

Quality of life

Health-related quality of life

EQ-5D[17]

Basic Needs

Housing type, stability including recent incarceration

Citation pending. Contact CNICS.

Social support

Multifactorial Assessment of Perceived Social Support-Short Form (MAPSS-SF)[18]

HIV-related stigma

HIV Stigma Mechanism Measure (adapted)[19]

Other

Family History (diabetes, heart disease, etc)

 

Health Care Utilization Data

  • CNICS captures health care utilization data from outpatient encounter/appointment systems and hospital systems as shown in the table below

Health Care Utilization Data

Outpatient Visit Data

HIV Primary Care

Subspecialty Care

Emergency Department

Telehealth

Outpatient Visit Appointment Data

Arrived

Bumped

Cancelled

No show

Pending

Hospitalization Data

Admission and discharge dates

Insurance Data

Public, Medicaid, Medicare, Ryan White, Private, Uninsured/self-pay

Genetic Data

  • CNICS has collected genetic data on over 10,000 participants using the Illumina LCG chip with 2.4 million common and rare variants
    • Genetic data were used to develop the Heat Map of Polygenic Risk Scores in the CNICS Cohort (European American and African American Sub-cohorts Combined) shown in the figure below

    The scores were generated using single nucleotide polymorphism-level effect estimates at multiple p-value cutoffs from previously published genome-wide association studies.  indicates associations significant at 10% false discovery rate. HDL, high density lipoprotein; LDL, low density lipoprotein; BMI, body mass index; CAD, coronary artery disease; MI, myocardial infarction.

    Chang H, Sewda A, Marquez-Luna C, White SR, Whitney BM, Williams-Nguyen J, Nance RM, Lee WJ, Kitahata MM, Saag MS, Willig A, Eron JJ, Mathews WC, Hunt PW, Moore RD, Webel A, Mayer KH, Delaney JAC, Crane PK, Crane HM, Hao K, and Peter I. Genetic Architecture of Cardiometabolic Risks in People Living with HIVBMC Medicine. In press

    Vital Status

    • CNICS sites maintain local death registries and collect death data from State Death Certificates and National Death Indexes to ensure complete ascertainment of death dates
    • Data regarding causes of death are obtained from the National Death Index (NDI)+, State Death Certificates, and medical record review (causes of death are unknown for approximately 35% of the CNICS cohort overall)

      Geographic Data

      • CNICS sites submit geographic information on participants’ place of residence in compliance with HIPAA requirements including aggregate geographic levels such as city, state, and 5-digit ZIP code 

      Biologic Specimens

      • The CNICS Specimen Core provides access to biologic specimens linked to patients’ comprehensive clinical data in the CNICS Research Platform. Information about specimens available in the CNICS Specimen Repository is provided here