These functions determine which items in a vector can be considered (the start of) a new episode, based on the argument episode_days. This can be used to determine clinical episodes for any epidemiological analysis. The get_episode() function returns the index number of the episode per group, while the is_new_episode() function returns values TRUE/FALSE to indicate whether an item in a vector is the start of a new episode.
Arguments
- x
vector of dates (class
DateorPOSIXt), will be sorted internally to determine episodes- episode_days
required episode length in days, can also be less than a day or
Inf, see Details- ...
ignored, only in place to allow future extensions
Details
Dates are first sorted from old to new. The oldest date will mark the start of the first episode. After this date, the next date will be marked that is at least episode_days days later than the start of the first episode. From that second marked date on, the next date will be marked that is at least episode_days days later than the start of the second episode which will be the start of the third episode, and so on. Before the vector is being returned, the original order will be restored.
The first_isolate() function is a wrapper around the is_new_episode() function, but is more efficient for data sets containing microorganism codes or names and allows for different isolate selection methods.
The dplyr package is not required for these functions to work, but these functions do support variable grouping and work conveniently inside dplyr verbs such as filter(), mutate() and summarise().
Examples
# `example_isolates` is a data set available in the AMR package.
# See ?example_isolates
df <- example_isolates[sample(seq_len(2000), size = 200), ]
get_episode(df$date, episode_days = 60) # indices
#> [1] 38 61 52 31 46 25 45 60 7 31 63 47 7 10 31 45 9 33 6 14 36 25 14 59 36
#> [26] 51 63 8 27 7 47 58 47 22 43 24 12 33 13 38 19 44 34 22 16 63 62 60 23 52
#> [51] 56 59 52 7 6 57 26 15 18 58 33 19 52 28 9 43 48 60 55 43 40 57 54 49 51
#> [76] 5 16 27 46 4 44 52 18 44 9 43 25 18 8 9 48 40 26 48 17 27 59 9 13 36
#> [101] 63 24 45 55 1 15 2 56 34 54 32 5 14 2 23 37 60 2 22 53 3 32 50 63 42
#> [126] 58 35 10 26 10 17 13 3 59 3 41 48 30 50 1 6 10 37 50 40 2 39 47 5 51
#> [151] 9 60 8 40 3 62 21 2 7 3 30 4 30 26 14 41 48 13 29 22 31 5 11 38 56
#> [176] 56 58 11 56 39 9 40 60 32 39 52 37 4 1 59 20 18 2 24 61 21 28 3 54 62
is_new_episode(df$date, episode_days = 60) # TRUE/FALSE
#> [1] FALSE TRUE FALSE FALSE FALSE TRUE FALSE FALSE FALSE TRUE TRUE FALSE
#> [13] FALSE TRUE FALSE TRUE FALSE TRUE TRUE FALSE FALSE FALSE FALSE FALSE
#> [25] FALSE FALSE FALSE FALSE TRUE TRUE FALSE FALSE FALSE FALSE TRUE FALSE
#> [37] TRUE FALSE TRUE TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
#> [49] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE TRUE FALSE FALSE TRUE
#> [61] FALSE TRUE TRUE TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE TRUE
#> [73] TRUE TRUE TRUE FALSE TRUE FALSE TRUE TRUE FALSE FALSE FALSE TRUE
#> [85] FALSE FALSE FALSE TRUE TRUE FALSE FALSE TRUE FALSE TRUE FALSE FALSE
#> [97] TRUE FALSE FALSE TRUE FALSE TRUE FALSE TRUE FALSE TRUE FALSE FALSE
#> [109] TRUE FALSE FALSE FALSE FALSE FALSE TRUE TRUE TRUE TRUE TRUE TRUE
#> [121] FALSE TRUE FALSE FALSE TRUE FALSE TRUE FALSE FALSE FALSE TRUE FALSE
#> [133] FALSE FALSE TRUE TRUE FALSE FALSE TRUE TRUE FALSE FALSE FALSE FALSE
#> [145] FALSE FALSE TRUE TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE TRUE
#> [157] TRUE FALSE FALSE FALSE TRUE FALSE FALSE FALSE TRUE FALSE FALSE FALSE
#> [169] TRUE FALSE FALSE TRUE FALSE FALSE TRUE FALSE FALSE TRUE FALSE FALSE
#> [181] TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE TRUE FALSE
#> [193] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
# filter on results from the third 60-day episode only, using base R
df[which(get_episode(df$date, 60) == 3), ]
#> # A tibble: 6 × 46
#> date patient age gender ward mo PEN OXA FLC AMX
#> <date> <chr> <dbl> <chr> <chr> <mo> <rsi> <rsi> <rsi> <rsi>
#> 1 2002-07-28 F54261 69 M Clinical B_STPHY_CONS R NA S NA
#> 2 2002-06-23 798871 82 M Clinical B_ENTRC_FCLS NA NA NA NA
#> 3 2002-06-06 24D393 20 F Clinical B_ESCHR_COLI R NA NA NA
#> 4 2002-06-06 24D393 20 F Clinical B_ESCHR_COLI R NA NA NA
#> 5 2002-07-30 218912 76 F ICU B_ESCHR_COLI R NA NA NA
#> 6 2002-07-15 426426 67 F ICU B_ESCHR_COLI R NA NA NA
#> # … with 36 more variables: AMC <rsi>, AMP <rsi>, TZP <rsi>, CZO <rsi>,
#> # FEP <rsi>, CXM <rsi>, FOX <rsi>, CTX <rsi>, CAZ <rsi>, CRO <rsi>,
#> # GEN <rsi>, TOB <rsi>, AMK <rsi>, KAN <rsi>, TMP <rsi>, SXT <rsi>,
#> # NIT <rsi>, FOS <rsi>, LNZ <rsi>, CIP <rsi>, MFX <rsi>, VAN <rsi>,
#> # TEC <rsi>, TCY <rsi>, TGC <rsi>, DOX <rsi>, ERY <rsi>, CLI <rsi>,
#> # AZM <rsi>, IPM <rsi>, MEM <rsi>, MTR <rsi>, CHL <rsi>, COL <rsi>,
#> # MUP <rsi>, RIF <rsi>
# the functions also work for less than a day, e.g. to include one per hour:
get_episode(c(
Sys.time(),
Sys.time() + 60 * 60
),
episode_days = 1 / 24
)
#> [1] 1 2
# \donttest{
if (require("dplyr")) {
# is_new_episode() can also be used in dplyr verbs to determine patient
# episodes based on any (combination of) grouping variables:
df %>%
mutate(condition = sample(
x = c("A", "B", "C"),
size = 200,
replace = TRUE
)) %>%
group_by(condition) %>%
mutate(new_episode = is_new_episode(date, 365)) %>%
select(patient, date, condition, new_episode)
}
#> # A tibble: 200 × 4
#> # Groups: condition [3]
#> patient date condition new_episode
#> <chr> <date> <chr> <lgl>
#> 1 0D7D34 2011-03-17 C TRUE
#> 2 D80438 2017-07-03 C FALSE
#> 3 590035 2015-02-18 A FALSE
#> 4 A86006 2009-08-09 A FALSE
#> 5 2D5F7B 2013-04-08 A FALSE
#> 6 7A1065 2008-02-11 A FALSE
#> 7 E4F322 2012-12-23 B FALSE
#> 8 D39422 2017-05-17 C TRUE
#> 9 C27336 2003-09-22 B FALSE
#> 10 006827 2009-07-24 A FALSE
#> # … with 190 more rows
if (require("dplyr")) {
df %>%
group_by(ward, patient) %>%
transmute(date,
patient,
new_index = get_episode(date, 60),
new_logical = is_new_episode(date, 60)
)
}
#> # A tibble: 200 × 5
#> # Groups: ward, patient [188]
#> ward date patient new_index new_logical
#> <chr> <date> <chr> <dbl> <lgl>
#> 1 ICU 2011-03-17 0D7D34 1 TRUE
#> 2 Clinical 2017-07-03 D80438 1 TRUE
#> 3 Clinical 2015-02-18 590035 1 TRUE
#> 4 Clinical 2009-08-09 A86006 1 TRUE
#> 5 ICU 2013-04-08 2D5F7B 1 TRUE
#> 6 ICU 2008-02-11 7A1065 1 TRUE
#> 7 ICU 2012-12-23 E4F322 1 TRUE
#> 8 Clinical 2017-05-17 D39422 1 TRUE
#> 9 ICU 2003-09-22 C27336 1 TRUE
#> 10 Clinical 2009-07-24 006827 1 TRUE
#> # … with 190 more rows
if (require("dplyr")) {
df %>%
group_by(ward) %>%
summarise(
n_patients = n_distinct(patient),
n_episodes_365 = sum(is_new_episode(date, episode_days = 365)),
n_episodes_60 = sum(is_new_episode(date, episode_days = 60)),
n_episodes_30 = sum(is_new_episode(date, episode_days = 30))
)
}
#> # A tibble: 3 × 5
#> ward n_patients n_episodes_365 n_episodes_60 n_episodes_30
#> <chr> <int> <int> <int> <int>
#> 1 Clinical 106 14 49 68
#> 2 ICU 69 13 40 52
#> 3 Outpatient 13 9 13 14
if (require("dplyr")) {
# grouping on patients and microorganisms leads to the same
# results as first_isolate() when using 'episode-based':
x <- df %>%
filter_first_isolate(
include_unknown = TRUE,
method = "episode-based"
)
y <- df %>%
group_by(patient, mo) %>%
filter(is_new_episode(date, 365)) %>%
ungroup()
identical(x, y)
}
#> Including isolates from ICU.
#> [1] FALSE
if (require("dplyr")) {
# but is_new_episode() has a lot more flexibility than first_isolate(),
# since you can now group on anything that seems relevant:
df %>%
group_by(patient, mo, ward) %>%
mutate(flag_episode = is_new_episode(date, 365)) %>%
select(group_vars(.), flag_episode)
}
#> # A tibble: 200 × 4
#> # Groups: patient, mo, ward [193]
#> patient mo ward flag_episode
#> <chr> <mo> <chr> <lgl>
#> 1 0D7D34 F_CANDD ICU TRUE
#> 2 D80438 B_CRYNB_STRT Clinical TRUE
#> 3 590035 B_ESCHR_COLI Clinical TRUE
#> 4 A86006 B_STPHY_CONS Clinical TRUE
#> 5 2D5F7B B_HMPHL_INFL ICU TRUE
#> 6 7A1065 B_ESCHR_COLI ICU TRUE
#> 7 E4F322 B_ENTRC ICU TRUE
#> 8 D39422 B_KLBSL_PNMN Clinical TRUE
#> 9 C27336 B_BCTRD_FRGL ICU TRUE
#> 10 006827 B_ENTRC_FCLS Clinical TRUE
#> # … with 190 more rows
# }