Looking for some inspiration for a message to start the new year, I pulled up my article from January 2020. There for perpetuity is a piece brimming with optimism for the design innovations of 2019, and how years of census planning had culminated in the start of 2020 Census operations in Alaska. The year 2020 would be one for taking advantage of decennial investments in communication, technology and enumerator manpower.

I did not have the prescience to mention that a pandemic would disrupt everything about the way we did business.

The experience of this year put two things in stark relief:

First, the degree to which the pandemic has impacted every dimension of American life – our economy, health, schools, public safety, food and housing security – would be quite literally immeasurable without the statistics produced by Federal agencies. As we faced the reality of curtailing in-person interviewing and production activities at the National Processing Center (NPC) in the interests of staff and public safety, we also knew that not collecting these data was not an option. We moved CAPI interviews to telephone and worked with the Regional Offices to tap their back-up inventory of respondent materials to get letters to respondents until NPC could set up a safe environment for its staff. Working with data from the Centers for Disease Control and Prevention and state agencies, our Field management has surgically opened and closed geographies for in-person interviewing throughout the year as health conditions dictated, demanding continuous communication with Census Bureau survey management teams and our sponsoring agencies.

Second, the pandemic has, perversely, sparked opportunities for innovation. Many of you stepped up in late March to contribute to the concept of a Household Pulse Survey to provide near real-time data on the impacts of the pandemic on key dimensions of American life. Within 35 days, the Federal statistical system conceived this survey and launched it into production. The first data files were available for use within a month, and they are released every two weeks. With counsel and support from the Office of Management and Budget, we have had the flexibility to adapt survey content and address evolving information needs. The Pulse also presented an opportunity to use a new software-as-service platform for rapid instrument development, and to contact respondents using SMS texts. Some of the established surveys also researching new approaches: the National Center for Health Statistics, for example, is piloting a survey for teens sampled from the NHIS roster; and conducting a follow-back on its 2019 households to supplement cases and mitigate lower response rates in 2020 due to limited in-person interviewing. There is renewed commitment to exploring the internet as a supplemental collection mode. The pandemic has introduced urgency into developing new approaches for virtual training for field staff, including for the upcoming New York City Housing Vacancy Survey where field representatives are given a degree of flexibility to sign up for a suite of training modules in a manner that works for their schedules.

The year 2021 will continue to be challenging, and we will work with each survey sponsor to keep planning and operations on track. In preparing 2020 and 2021 data products, there will be ample work needed to understand the impacts that operational disruptions have had on data quality. The Census Bureau and its Federal statistical partners remain committed to transparency in this process, and ensuring users understand the limitations in the data due to the pandemic.

Our challenge going into this year is to persevere in the continuity of operations, adapting as necessary – and exercising the innovative spirit that has sparked in this unprecedented time. As we get closer to the end of this experience this year, please accept our appreciation for the resourcefulness and flexibility on the part of you and your agency team.

Surveying America’s Health Care Workers During the COVID-19 Pandemic: What Have We Learned So Far

by Lucinda Dalzell, Survey Director for National Ambulatory Medical Care Surveys, Demographic Programs Directorate, U.S. Census Bureau

The Census Bureau is the data collection agent for the National Ambulatory Medical Care Survey (NAMCS) and the National Hospital Ambulatory Medical Care Survey (NHAMCS), sponsored by the National Center for Health Statistics (NCHS). Respondents for both surveys are on the front lines of the COVID-19 pandemic, with NAMCS collecting data about patient visits to doctor’s offices and community health centers and NHAMCS collecting data about visits to hospital emergency departments.

How has the pandemic impacted 2020 data collection efforts, and how have NCHS and the Census Bureau responded to the situation? More importantly, what have we learned from this experience that should or should not be applied to similar situations in the future? Our current thoughts are laid out below.

The 2020 data collection for both surveys was originally scheduled to begin in the spring of 2020. We knew from news reports that while some areas were initially very hard hit with a surge in COVID-19 cases, health care workers all over the country were preparing for the virus to impact their area, even if it had not yet arrived there. Because of this, it was quickly apparent that we should delay the start of data collection across the entire country, which we did. What was less apparent was when to start. Unlike a natural disaster, where the worst happens and then the situation slowly improves, with COVID-19, the specific geographies hardest hit kept changing, with little overall relief as the weather warmed. We did start 2020 data collection in June of 2020, with a longer delay in those areas most impacted during the spring.

Thinking through how to apply this experience in the future, I think we would want to take a similar approach. The trends in cases of infection were becoming apparent to everyone globally in real-time, so on the survey management end, we forecasted the future trajectory and its impact based on these reports and had to accept the limitations of the knowledge available. We also had to be prepared to adapt our data collection strategies, which could potentially include ceasing data collection for a time as the weather got colder. Fortunately, to date, we haven’t had to do this, and a lot of that is due to the other strategies we employed below.
Our survey team appreciated strong lines of communication from the Field Representatives (FRs), through the Regional Offices (ROs), FLD HQ, through ADDP-SO and over to NCHS. When we initially began 2020 data collection, after a few weeks, FRs reported back to us that hospitals in geographies we originally had held off on had begun to reopen to non-patients and could potentially participate in the survey, while other geographies not on our delayed start list had recently experienced a surge and were not allowing non-patients to come on-site. With this information, NCHS made the decision to begin data collection everywhere and was open to the idea of brief extensions in closeout dates for certain cases especially affected by COVID-19.
Another important innovation we made early in our data collection efforts was to establish a living document on the processes by which ROs and FRs could arrange for remote abstraction at the ROs. Abstraction is the phase of data collection whereby FRs are provided with information about a sample of records of patient visits during the reporting period and they enter the relevant fields into the survey instrument. With physician’s offices and hospitals limiting on-site access, alternative options were needed. NCHS had previously approved some specific remote abstraction modes. As respondents made additional requests for other IT software options for data sharing, NCHS reviewed these for confidentiality and security concerns and helped Census implement several new options.
While the remote abstraction process worked sufficiently well for NHAMCS, it was more difficult for NAMCS with its much larger caseload and respondent’s whose businesses have fewer resources. For these reasons, after the first quarter of NAMCS, NCHS made the decision to cease the abstraction phase of the survey and only conduct the true “interview” portions of the survey for the physician cases (we are still abstracting community health centers). While not every survey would be able to drop the most burdensome portions of their data collection, it so far has been a successful decision to have made on NAMCS. While we can’t necessarily attribute this solely to dropping abstraction, the second quarter of NAMCS closed out 4 weeks sooner and with a response rate 10 percentage points higher on its interview portions than the first quarter.
NAMCS and NHAMCS have a folder of informational materials which the FRs traditionally provide to respondents in person. With limited personal visits as an option, we recreated this folder in electronic format so that it could be provided over e-mail. This has been an advance in our outreach efforts that we will likely keep even after the pandemic abates.
FR morale was and remains a strong concern of the survey team. On the one hand, data on health care has never been more critical. On the other hand, the FRs are faced with trying to gain cooperation from health care workers in a very stressful situation, while at the same time losing income from the limited ability to travel. To help address this, NCHS approved some monetary awards where possible. HQ also reached out to individual FRs in the form of thank you notes and personal phone calls to acknowledge the significant efforts they were going through to collect this vital data. The FRs were clear they very much appreciated all these signs of appreciation. We should look to expanding such initiatives in the future.

Data collection for the third quarter of NAMCS and third trimester of NHAMCS began in December 2020. For NAMCS, COVID-19-related questions have been added. Specifically, those questions ask doctor's offices and community health centers about their experiences during the last 3 months regarding: 1) shortages of personal protective equipment (PPE), 2) ability to test patients for COVID-19, 3) did the office need to turn away or refer elsewhere patients diagnosed with COVID-19, 4) have any health care providers tested positive with COVID-19, and 5) use of telemedicine/telehealth technologies. Such questions will add additional value to data on health care during the COVID-19 pandemic, as there is limited information on how this virus has impacted such health care settings.

As a survey team, we will continue to reassess our current lessons learned and gather additional lessons. This will help NAMCS, NHAMCS, and other surveys determine what recent modifications should remain long-term, as well as better prepare us should we unfortunately find ourselves in a similar situation in the future.

2021 New York City Housing and Vacancy Survey

by Tamara Cole, Survey Director for Housing Surveys, Demographic Programs Directorate, U.S. Census Bureau

Data collection began on February 1 for the 2021 New York City Housing and Vacancy Survey (NYCHVS). The NYCHVS is conducted every three years and is funded by the New York City Department of Housing Preservation and Development (HPD). The survey is used to determine the vacancy rate for New York City's rental stock and to provide a measure of housing conditions, as required by law. New York City also uses the data to measure the quality and quantity of housing and the demographic characteristics of the city's residents.

The Census Bureau has conducted this survey for New York City since 1965. The NYCHVS has been redesigned for 2021, which was the first full redesign since 1990. At the conclusion of the 2017 NYCHVS, a comprehensive review was conducted to identify opportunities to increase the value of the survey for current and future generations of users. This review was completed by Sponsor staff at HPD and incorporated feedback from conversations with more than two dozen organizations, including NYCHVS users at other NYC agencies, non-profits, and academic institutions as well as ongoing conversations with multiple divisions within the United States Census Bureau.

Some of the key findings and recommendations include the following:

Researchers rely on NYCHVS data to understand the current state of housing and how key aspects of the stock have changed over time.
The NYCHVS is the primary source of comprehensive information on rent regulated units; the survey should continue to focus on this subset of the housing stock and resident population and incorporate additional information where possible to understand how this segment of New York City is changing.
The NYCHVS facilitates comparisons among different subsets of the population, including by race/ethnicity, nativity, age, and household composition; the survey should collect information to facilitate analysis of issues related to fair housing.
The NYCHVS is a source of information that has, and should, support decision making in various policy domains beyond housing; to the extent possible, the survey should expand the types of questions it collects to support more coordinated policymaking, including information on household costs, ability to meet critical expenses, and debt/assets.

For 2021 specifically, we have a unique opportunity to provide timely data on the effect of the COVID-19 pandemic on New York City. The survey will collect information about respondent’s experience during the pandemic (including their changes in household, employment, health, and financial well-being).

This survey provides data on the characteristics of housing units, their occupants, buildings and neighborhoods. Data are available, among other items, on household income, poverty income levels, monthly rent, rent/income ratio, rent received from government programs, monthly mortgage payments, utility costs, routine maintenance fees, and real estate taxes. Data are shown for education and labor force characteristics for all household members 15 years and older. Data are also presented on the condition of the building and neighborhood, the occupancy status of the unit, the year the building was built, whether the unit is owner or renter-occupied, condominium or cooperative status of the building, and wheelchair accessibility. Additional data focus on number of stories, units, and rooms in the unit; the type of heating fuel, maintenance indicators such as cracks or holes in walls or floors; and the occurrence of heating equipment breakdowns.

Approximately 12,000 units throughout the city are included as a representative sample of the housing in the five boroughs of New York City. Results from the 2021 NYCHVS will be available early next year.

You can find more information and statistics online at https://www1.nyc.gov/site/hpd/about/research.page or https://www.census.gov/programs-surveys/nychvs.html

Mobile Case Management's Migration Journey

by Christopher Kamsler - Lead Architect Mobile Case Management, U.S. Census Bureau

The Census Bureau’s Mobile Case Management (MCM) system achieved a major milestone in December 2020, with the migration of the final surveys to its environment. MCM is one of the control systems the Census Bureau uses to transfer survey instruments and data between Census Headquarters, the Regional Offices, and the Field Representatives (FRs). MCM is used by the FRs to receive and manage case workload for the various surveys assigned to them.

MCM replaces the Bureau’s Laptop Case Management (LCM), which had become obsolete. MCM has several benefits. First, it was built with flexibilities to allow it to grow with current technology trends and to be able to support Census Bureau initiatives for years to come. For example, it is flexible enough to allow it to interface with different operational control systems (OCS), such as Field OCS for Decennial, the Regional Office Survey Control System (ROSCO) for current surveys and the Post-Enumeration Survey, the upcoming DICE or our Data Ingest and Collection for the Enterprise System, and DOCS. In addition, MCM allows for cases from multiple surveys to display on the same home screen, whereas LCM had a different instance for each survey. MCM also provides map and calendar functionality on the home screen to help FRs plan how they will perform their work. All these features help FRs better manage their workload, as many of them work on multiple surveys.

The migration from LCM to MCM has taken a number of years to plan and execute across the full suite of surveys.

MCM first went live in December 2016 for the American Community Survey. Lessons from that experience led to architectural improvements, including migrating the user base over to a new SQLite client database system that introduced improved stability and performance.

Since then the MCM team steadily converted the remainder of the surveys dependent on LCM: