COVID Literature since January 1, 2020
Mohleen Kang, MD; Saumya Gurbani; Jordan A. Kempker, MD, MSc
2020-06-09
Introduction
The rapid pace of publicatins regarding COVID-19 has been staggering. Here we begin to describe the prolieferaion of the published literature.
We utilized the National Library of Medicines search string:
> ((wuhan[All Fields] AND (“coronavirus”[MeSH Terms] OR “coronavirus”[All Fields])) AND 2019/12[PDAT] : 2030[PDAT]) OR 2019-nCoV[All Fields] OR 2019nCoV[All Fields] OR COVID-19[All Fields] OR SARS-CoV-2[All Fields]
Daily new cases of COVID-19 were extracted from the website Our World in Data
Dataset Construction
Exclusions are those published before january 1, 2020 and those published after the last Monday (June 8) before the search query was run. Select the code button in this section to see the R code for dataset construction.
setwd("C:/Users/jkempke/Box Sync/Side Projects/COVID/Global Publications/Data")
load("covid_pubmed_20200609_1000.RData")
# Some initial cleaning
pub_db1 <- pub_db %>% filter(DEP > "2020-01-01") %>% filter(DEP < "2020-06-08") %>%
select(LA, PST, PL, DEP, PMID, PT, PT2, TA, ) %>% mutate(day = weekdays(DEP),
LA = droplevels(LA), PL = droplevels(PL), week = isoweek(DEP), week.cat = as.factor(week),
LA = if_else(LA == "chi", "Chinese", if_else(LA == "dut", "Dutch", if_else(LA ==
"eng", "English", if_else(LA == "fre", "French", if_else(LA == "ger", "German",
if_else(LA == "hun", "Hungarian", if_else(LA == "ita", "Italian", if_else(LA ==
"pol", "Polish", if_else(LA == "por", "Portuguese", if_else(LA == "rus",
"Russian", if_else(LA == "spa", "Spanish", if_else(LA == "swe", "Swedish",
NA_character_)))))))))))), PT = as.character(PT), PT2 = as.character(PT2),
PT.adj = if_else(PT2 == "", PT, if_else(is.na(PT2) == F, PT2, NA_character_))) %>%
mutate(PT.adj = if_else(PT == "Case Reports", PT, PT.adj))
# Dataframe for exaing freuqncies by Journal
pub_db.journal <- pub_db1 %>% group_by(TA) %>% summarise(N = n())
pub_db.journal <- pub_db.journal %>% mutate(Total.N = sum(N), Percent = round(100 *
N/Total.N, digits = 1)) %>% dplyr::select(-Total.N) %>% filter(Percent > 0.4) %>%
arrange(desc(Percent))
# Data frame for associating weeks with the date of its corersponding Monday
date <- seq(as.Date("2020/01/06"), by = "weeks", length.out = 22)
mondays <- as.data.frame(date)
mondays <- mondays %>% mutate(day = "Monday", week = isoweek(date))
# Create time series for graphing, 1 aggregated by day another by week
pub_db.day <- pub_db1 %>% group_by(DEP) %>% summarise(N = n()) %>% mutate(Cum.N = cumsum(N),
day = weekdays(DEP))
pub_db.week <- pub_db1 %>% group_by(week) %>% summarise(N = n()) %>% mutate(Cum.N = cumsum(N))
# Xcluded first two weeks of january for low pub counts as outliers.
pub_db.week.trimmed <- filter(pub_db.week, week > 3)
pub_db.weekts <- left_join(pub_db.week, mondays, by = "week")
# now import all of the global COVID daily new case numbers and merge
global.covid <- read.csv("owid-covid-data.csv")
global.covid <- global.covid %>% dplyr::select(date, new_cases) %>% mutate(date = as.character(date),
Date = as_date(date), week = isoweek(Date)) %>% filter(Date > "2020-01-01") %>%
filter(Date < "2020-06-08")
global.covid.day <- global.covid %>% group_by(Date) %>% summarise(N = sum(new_cases)) %>%
mutate(Cum.N = cumsum(N), day = weekdays(Date))
global.covid.week <- global.covid %>% group_by(week) %>% summarise(Cases = sum(new_cases)) %>%
mutate(Cum.Cases = cumsum(Cases))
weekts <- left_join(pub_db.weekts, global.covid.week, by = "week")Time Series
- General statistics:
- Weekly mean (SD) = 803.3181818 (778.402845)
- 10th percentile = 17.2
- 25th percentile = 90
- 50th percentile = 473.5
- 75th percentile = 1626.5
- 90th percentile = 1771.3
- Removing the first 2 weeks of January 2020:
- Weekly mean (SD) = 883.45 (771.5836398)
- 10th percentile = 74.8
- 25th percentile = 159
- 50th percentile = 748.5
- 75th percentile = 1693.5
- 90th percentile = 1775.8
ggplot(pub_db.day, aes(x = DEP, fill = day)) + geom_bar(aes(y = N), stat = "identity",
color = "black") + theme_bw() + labs(x = "Electronic Publication Date", y = "Number of Publications") +
scale_x_date(breaks = "1 week", date_labels = "%b %d") + theme(axis.title = element_text(size = 14),
axis.text.x = element_text(angle = 60, hjust = 1, size = 12), axis.text.y = element_text(size = 12))
ggplot(pub_db.weekts, aes(x = date, y = N)) + geom_bar(stat = "identity", color = "black") +
theme_bw() + geom_text(aes(label = N), nudge_y = 500) + scale_x_date(breaks = "1 week",
date_labels = "%b %d") + labs(x = "Week of Electronic Publication Date", y = "Number of Publications") +
theme(axis.title = element_text(size = 14), axis.text.x = element_text(angle = 60,
hjust = 1, size = 12), axis.text.y = element_text(size = 12))
ggplot(weekts, aes(x = date)) + geom_bar(aes(y = N), stat = "identity", color = "black",
fill = "light blue") + geom_line(aes(y = Cases/1000), size = 1) + theme_bw() +
geom_text(aes(y = N, label = N), nudge_y = 100) + scale_x_date(breaks = "1 week",
date_labels = "%b %d") + scale_y_continuous(name = "Number of Publications",
sec.axis = sec_axis(trans = ~. + 0, name = "Number of Global Cases (in thousands)")) +
labs(x = "Monday Date") + theme(axis.title = element_text(size = 14), axis.text.x = element_text(angle = 60,
hjust = 1, size = 12), axis.text.y = element_text(size = 12))
ggplot(pub_db.weekts, aes(x = date, y = Cum.N)) + geom_bar(stat = "identity", color = "black") +
geom_text(aes(label = Cum.N), nudge_y = 1000) + theme_bw() + scale_x_date(breaks = "1 week",
date_labels = "%b %d") + labs(x = "Week of Electronic Publication Date", y = "Number of Publications") +
theme(axis.title = element_text(size = 14), axis.text.x = element_text(angle = 60,
hjust = 1, size = 12), axis.text.y = element_text(size = 12))
Publication Characteristics
compare.pubs1 <- compareGroups(~PL + LA + PST + day + week.cat, data = pub_db1, max.xlev = 60)
compare.pubs1.table <- createTable(compare.pubs1, hide.no = "no")
export2md(compare.pubs1.table, loc = "left", size = "large")| [ALL] | N | |
|---|---|---|
| N=17673 | ||
| PL: | 17672 | |
| Argentina | 3 (0.02%) | |
| Australia | 204 (1.15%) | |
| Austria | 26 (0.15%) | |
| Belgium | 1 (0.01%) | |
| Bosnia and Herzegovina | 2 (0.01%) | |
| Brazil | 144 (0.81%) | |
| Canada | 350 (1.98%) | |
| Chile | 8 (0.05%) | |
| China | 281 (1.59%) | |
| China (Republic : 1949- ) | 7 (0.04%) | |
| Czech Republic | 1 (0.01%) | |
| Denmark | 80 (0.45%) | |
| Egypt | 5 (0.03%) | |
| England | 4849 (27.4%) | |
| France | 283 (1.60%) | |
| Germany | 565 (3.20%) | |
| Greece | 12 (0.07%) | |
| Hungary | 15 (0.08%) | |
| India | 152 (0.86%) | |
| Iran | 57 (0.32%) | |
| Ireland | 275 (1.56%) | |
| Italy | 289 (1.64%) | |
| Japan | 71 (0.40%) | |
| Korea (South) | 128 (0.72%) | |
| Malaysia | 4 (0.02%) | |
| Mexico | 13 (0.07%) | |
| Nepal | 11 (0.06%) | |
| Netherlands | 1470 (8.32%) | |
| New Zealand | 48 (0.27%) | |
| Norway | 37 (0.21%) | |
| Oman | 1 (0.01%) | |
| Poland | 49 (0.28%) | |
| Portugal | 17 (0.10%) | |
| Romania | 3 (0.02%) | |
| Russia (Federation) | 1 (0.01%) | |
| Saudi Arabia | 5 (0.03%) | |
| Scotland | 38 (0.22%) | |
| Singapore | 74 (0.42%) | |
| South Africa | 10 (0.06%) | |
| Spain | 241 (1.36%) | |
| Sweden | 21 (0.12%) | |
| Switzerland | 634 (3.59%) | |
| Thailand | 2 (0.01%) | |
| Turkey | 71 (0.40%) | |
| United Arab Emirates | 26 (0.15%) | |
| United States | 7088 (40.1%) | |
| LA: | 17673 | |
| Chinese | 103 (0.58%) | |
| Dutch | 14 (0.08%) | |
| English | 17250 (97.6%) | |
| French | 92 (0.52%) | |
| German | 59 (0.33%) | |
| Hungarian | 13 (0.07%) | |
| Italian | 2 (0.01%) | |
| Polish | 1 (0.01%) | |
| Portuguese | 23 (0.13%) | |
| Russian | 2 (0.01%) | |
| Spanish | 106 (0.60%) | |
| Swedish | 8 (0.05%) | |
| PST: | 17673 | |
| aheadofprint | 11435 (64.7%) | |
| epublish | 2899 (16.4%) | |
| ppublish | 3339 (18.9%) | |
| day: | 17673 | |
| Friday | 3498 (19.8%) | |
| Monday | 2718 (15.4%) | |
| Saturday | 1388 (7.85%) | |
| Sunday | 745 (4.22%) | |
| Thursday | 3256 (18.4%) | |
| Tuesday | 2871 (16.2%) | |
| Wednesday | 3197 (18.1%) | |
| week.cat: | 17673 | |
| 2 | 3 (0.02%) | |
| 3 | 1 (0.01%) | |
| 4 | 15 (0.08%) | |
| 5 | 37 (0.21%) | |
| 6 | 79 (0.45%) | |
| 7 | 80 (0.45%) | |
| 8 | 120 (0.68%) | |
| 9 | 172 (0.97%) | |
| 10 | 182 (1.03%) | |
| 11 | 234 (1.32%) | |
| 12 | 400 (2.26%) | |
| 13 | 547 (3.10%) | |
| 14 | 950 (5.38%) | |
| 15 | 1107 (6.26%) | |
| 16 | 1306 (7.39%) | |
| 17 | 1684 (9.53%) | |
| 18 | 1775 (10.0%) | |
| 19 | 1722 (9.74%) | |
| 20 | 2284 (12.9%) | |
| 21 | 1783 (10.1%) | |
| 22 | 1738 (9.83%) | |
| 23 | 1454 (8.23%) |
Journal’s contriburinting to at least 0.5% of total publications
knitr::kable(pub_db.journal, format = "html")| TA | N | Percent |
|---|---|---|
| BMJ | 495 | 2.8 |
| J Med Virol | 344 | 1.9 |
| medRxiv | 241 | 1.4 |
| Clin Infect Dis | 169 | 1.0 |
| J Infect | 175 | 1.0 |
| Lancet | 183 | 1.0 |
| N Engl J Med | 177 | 1.0 |
| Dermatol Ther | 154 | 0.9 |
| Int J Infect Dis | 149 | 0.8 |
| JAMA | 139 | 0.8 |
| bioRxiv | 129 | 0.7 |
| Int J Environ Res Public Health | 117 | 0.7 |
| Infect Control Hosp Epidemiol | 113 | 0.6 |
| J Am Acad Dermatol | 112 | 0.6 |
| Lancet Infect Dis | 104 | 0.6 |
| Med Hypotheses | 112 | 0.6 |
| Sci Total Environ | 102 | 0.6 |
| Travel Med Infect Dis | 109 | 0.6 |
| Ann Intern Med | 83 | 0.5 |
| Brain Behav Immun | 89 | 0.5 |
| Crit Care | 97 | 0.5 |
| Head Neck | 85 | 0.5 |
| J Clin Virol | 84 | 0.5 |
| J Eur Acad Dermatol Venereol | 88 | 0.5 |
| Psychiatry Res | 82 | 0.5 |