Skip to content

Commit e7736fa

Browse files
authored
Merge pull request #315 from cmu-delphi/docs/add-nchs
Release NCHS Mortality
2 parents 58fc8cb + 83fe21f commit e7736fa

File tree

1 file changed

+101
-0
lines changed

1 file changed

+101
-0
lines changed
Lines changed: 101 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,101 @@
1+
---
2+
title: NCHS Mortality Data
3+
parent: Data Sources and Signals
4+
grand_parent: COVIDcast Epidata API
5+
---
6+
7+
# NCHS Mortality Data
8+
{: .no_toc}
9+
10+
* **Source name:** `nchs-mortality`
11+
* **First issued:** Epiweek 50 2020 (6-12 December 2020)
12+
* **Number of data revisions since 19 May 2020:** 0
13+
* **Date of last change:** Never
14+
* **Available for:** state (see [geography coding docs](../covidcast_geography.md))
15+
* **License:** [NCHS Data Use Agreement](https://www.cdc.gov/nchs/data_access/restrictions.htm)
16+
17+
This data source of national provisional death counts is based on death
18+
certificate data received and coded by the National Center for Health Statistics
19+
[(NCHS)](https://www.cdc.gov/nchs/nvss/vsrr/COVID19/index.htm).
20+
21+
| Signal | Description |
22+
| --- | --- |
23+
| `deaths_covid_incidence_num` | Number of weekly new deaths with confirmed or presumed COVID-19 |
24+
| `deaths_covid_incidence_prop` | Number of weekly new deaths with confirmed or presumed COVID-19, per 100,000 population |
25+
| `deaths_allcause_incidence_num` | Number of weekly new deaths from all causes |
26+
| `deaths_allcause_incidence_prop` | Number of weekly new deaths from all causes, per 100,000 population |
27+
| `deaths_flu_incidence_num` | Number of weekly new deaths involving Influenza and at least one of (Pneumonia, COVID-19)|
28+
| `deaths_flu_incidence_prop` | Number of weekly new deaths involving Influenza and at least one of (Pneumonia, COVID-19), per 100,000 population |
29+
| `deaths_pneumonia_notflu_incidence_num` | Number of weekly new deaths involving Pneumonia, excluding Influenza deaths |
30+
| `deaths_pneumonia_notflu_incidence_prop` | Number of weekly new deaths involving Pneumonia, excluding Influenza deaths, per 100,000 population |
31+
| `deaths_covid_and_pneumonia_notflu_incidence_num`| Number of weekly new deaths involving COVID-19 and Pneumonia, excluding Influenza |
32+
| `deaths_covid_and_pneumonia_notflu_incidence_prop`| Number of weekly new deaths involving COVID-19 and Pneumonia, excluding Influenza, per 100,000 population |
33+
|`deaths_pneumonia_or_flu_or_covid_incidence_num`| Number of weekly new deaths involving Pneumonia, Influenza, or COVID-19|
34+
|`deaths_pneumonia_or_flu_or_covid_incidence_prop`| Number of weekly new deaths involving Pneumonia, Influenza, or COVID-19, per 100,000 population|
35+
|`deaths_percent_of_expected`| Number of weekly new deaths for all causes in 2020 compared to the average number across the same week in 2017–2019|
36+
37+
These signals are taken directly from [Table
38+
1](https://www.cdc.gov/nchs/nvss/vsrr/COVID19/index.htm) without
39+
changes. National provisional death counts include deaths occurring within the
40+
50 states and the District of Columbia that have been received and coded as of
41+
the date specified during a given time period. The deaths are classified based
42+
on a new ICD-10 code. (Note that the classification is based on all the codes on
43+
the death certificate, not just the primary cause of death). The codes that are
44+
considered for each signals are described in detail
45+
[here](https://github.com/cmu-delphi/covidcast-indicators/blob/nchs_mortality/nchs_mortality/DETAILS.md#metrics-level-1-m1). We
46+
export the state-level data as-is in a weekly format.
47+
48+
## Table of contents
49+
{: .no_toc .text-delta}
50+
51+
1. TOC
52+
{:toc}
53+
54+
## Geographical Exceptions
55+
56+
New York City is listed as its own region in the NCHS Mortality data, but
57+
we don't consider NYC separately. The death counts for NYC are included in New
58+
York State in our reports.
59+
60+
## Report Using Epiweeks
61+
62+
We report the NCHS Mortality data in a weekly format (`time_type=week` \&
63+
`time_value=\{YYYYWW\}`, where `YYYYWW` refers to an epiweek). The CDC defines
64+
the [epiweek](https://wwwn.cdc.gov/nndss/document/MMWR_Week_overview.pdf) as
65+
seven days, from Sunday to Saturday. We check the week-ending dates provided in
66+
the NCHS morality data and use Python package
67+
[epiweeks](https://pypi.org/project/epiweeks/) to convert them into epiweek
68+
format.
69+
70+
## Missingness
71+
72+
NCHS suppresses some data to protect individual privacy and avoid publishing
73+
low-confidence figures. This includes data for jurisdictions where counts are
74+
between 1 and 9, and data for weeks where the counts are less than 50% of the
75+
expected number, since these provisional counts are highly incomplete and
76+
potentially misleading.
77+
78+
## Lag and Backfill
79+
80+
There is a lag in time between when the death occurred and when the death
81+
certificate is completed, submitted to NCHS, and processed for reporting
82+
purposes. The death counts for earlier weeks are continually revised and may
83+
increase or decrease as new and updated death certificate data are received from
84+
the states by NCHS. This delay can range from 1 to 8 weeks or even more.
85+
Some states report deaths on a daily basis, while other states report deaths weekly
86+
or monthly. State vital record reporting may also be affected or delayed by
87+
COVID-19 related response activities which make death counts not comparable
88+
across states. We check for updates reported by NCHS every weekday but will
89+
report the signals weekly (on Monday).
90+
91+
## Source and Licensing
92+
93+
This data was originally published by the National Center for Health Statistics,
94+
and is made available here as a convenience to the forecasting community under
95+
the terms of the original license. The NCHS places restrictions on how this
96+
dataset may be used: you may not attempt to identify any individual included in
97+
the data, whether by itself or through linking to other
98+
individually=identifiable data; you may only use the dataset for statistical
99+
reporting and analysis. The full text of the [NCHS Data Use
100+
Agreement](https://www.cdc.gov/nchs/data_access/restrictions.htm) is available
101+
from their website.

0 commit comments

Comments
 (0)