The health and safety of incarcerated persons and correctional personnel have been prominent in the U.S. news media discourse during the COVID-19 pandemic. Examining changing attitudes toward the health of the incarcerated population is imperative to better assess the extent to which the general public favors criminal justice reform. However, existing natural language processing lexicons that underlie current sentiment analysis (SA) algorithms may not perform adequately on news articles related to criminal justice due to contextual complexities. News discourse during the pandemic has highlighted the need for a novel SA lexicon and algorithm (i.e., an SA package) tailored for examining public health policy in the context of the criminal justice system. We analyzed the performance of existing SA packages on a corpus of news articles at the intersection of COVID-19 and criminal justice collected from state-level outlets between January and May 2020. Our results demonstrated that sentence sentiment scores provided by three popular SA packages can differ considerably from manually-curated ratings. This dissimilarity was especially pronounced when the text was more polarized, whether negatively or positively. A randomly selected set of 1,000 manually scored sentences, and the corresponding binary document term matrices, were used to train two new sentiment prediction algorithms (i.e., linear regression and random forest regression) to verify the performance of the manually-curated ratings. By better accounting for the unique context in which incarceration-related terminologies are used in news media, both of our proposed models outperformed all existing SA packages considered for comparison. Our findings suggest that there is a need to develop a novel lexicon, and potentially an accompanying algorithm, for analysis of text related to public health within the criminal justice system, as well as criminal justice more broadly.

Importance

Adults with Alzheimer disease and related dementias (ADRD) are particularly vulnerable to the direct and indirect effects of the COVID-19 pandemic. Deaths associated with ADRD increased substantially in pandemic year 1. It is unclear whether mortality associated with ADRD declined when better prevention strategies, testing, and vaccines became widely available in year 2.

Objective

To compare pandemic-era excess deaths associated with ADRD between year 1 and year 2 overall and by age, sex, race and ethnicity, and place of death.

Design, setting, and participants

This time series analysis used all death certificates of US decedents 65 years and older with ADRD as an underlying or contributing cause of death from January 2014 through February 2022.

Exposure

COVID-19 pandemic era.

Main outcomes and measures

Pandemic-era excess deaths associated with ADRD were defined as the difference between deaths with ADRD as an underlying or contributing cause observed from March 2020 to February 2021 (year 1) and March 2021 to February 2022 (year 2) compared with expected deaths during this period. Expected deaths were estimated using data from January 2014 to February 2020 fitted with autoregressive integrated moving average models.

Results

Overall, 2 334 101 death certificates were analyzed. A total of 94 688 (95% prediction interval [PI], 84 192-104 890) pandemic-era excess deaths with ADRD were estimated in year 1 and 21 586 (95% PI, 10 631-32 450) in year 2. Declines in ADRD-related deaths in year 2 were substantial for every age, sex, and racial and ethnic group evaluated. Pandemic-era ADRD-related excess deaths declined among nursing home/long-term care residents (from 34 259 [95% PI, 25 819-42 677] in year 1 to -22 050 [95% PI, -30 765 to -13 273] in year 2), but excess deaths at home remained high (from 34 487 [95% PI, 32 815-36 142] in year 1 to 28 804 [95% PI, 27 067-30 571] in year 2).

Conclusions and relevance

This study found that large increases in mortality with ADRD as an underlying or contributing cause of death occurred in COVID-19 pandemic year 1 but were largely mitigated in pandemic year 2. The most pronounced declines were observed for deaths in nursing home/long-term care settings. Conversely, excess deaths at home and in medical facilities remained high in year 2.