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
Date
orPOSIXt
), 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$date, episode_days = 60) # indices
#> [1] 23 4 57 10 24 33 54 46 15 13 52 3 46 30 17 43 31 50 27 6 44 18 9 40 36
#> [26] 5 32 49 49 9 23 27 8 8 22 53 26 58 1 59 49 59 5 35 22 11 43 60 54 13
#> [51] 51 26 18 43 2 22 47 28 36 22 15 1 40 23 41 34 60 32 45 9 51 21 50 24 35
#> [76] 42 26 29 58 31 13 59 4 55 38 10 19 54 29 7 53 50 55 15 40 4 46 55 61 9
#> [101] 14 43 54 45 54 10 23 55 18 48 9 33 29 7 6 41 39 28 59 7 62 6 41 57 32
#> [126] 60 41 35 37 15 45 49 24 36 10 25 22 20 52 14 62 8 6 42 60 12 41 14 32 21
#> [151] 59 58 9 13 13 21 36 3 49 12 1 9 57 3 17 23 14 13 62 56 57 40 6 13 15
#> [176] 19 57 24 47 59 20 61 46 54 58 48 30 39 4 32 42 25 4 22 20 56 3 16 44 11
is_new_episode(df$date, episode_days = 60) # TRUE/FALSE
#> [1] TRUE TRUE FALSE TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
#> [13] TRUE FALSE FALSE FALSE FALSE FALSE TRUE FALSE TRUE FALSE FALSE FALSE
#> [25] FALSE FALSE FALSE FALSE TRUE TRUE FALSE FALSE TRUE FALSE FALSE TRUE
#> [37] TRUE FALSE TRUE FALSE FALSE TRUE TRUE FALSE FALSE FALSE FALSE FALSE
#> [49] FALSE TRUE TRUE FALSE FALSE FALSE TRUE FALSE FALSE FALSE FALSE FALSE
#> [61] TRUE FALSE TRUE FALSE TRUE TRUE TRUE FALSE FALSE FALSE FALSE TRUE
#> [73] TRUE FALSE FALSE FALSE FALSE FALSE FALSE TRUE FALSE FALSE FALSE FALSE
#> [85] TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE TRUE FALSE FALSE FALSE
#> [97] FALSE FALSE TRUE FALSE FALSE TRUE TRUE TRUE FALSE FALSE FALSE FALSE
#> [109] TRUE TRUE FALSE TRUE TRUE TRUE TRUE FALSE TRUE TRUE FALSE FALSE
#> [121] FALSE FALSE FALSE TRUE FALSE FALSE FALSE TRUE TRUE FALSE FALSE FALSE
#> [133] TRUE FALSE FALSE TRUE FALSE FALSE TRUE FALSE TRUE FALSE FALSE TRUE
#> [145] FALSE FALSE FALSE FALSE TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
#> [157] TRUE FALSE FALSE TRUE FALSE FALSE FALSE FALSE TRUE FALSE TRUE FALSE
#> [169] FALSE TRUE FALSE FALSE FALSE FALSE FALSE TRUE FALSE FALSE TRUE FALSE
#> [181] TRUE FALSE FALSE FALSE TRUE FALSE TRUE FALSE FALSE FALSE FALSE FALSE
#> [193] FALSE TRUE FALSE FALSE TRUE TRUE FALSE TRUE
# filter on results from the third 60-day episode only, using base R
df[which(get_episode(df$date, 60) == 3), ]
#> # A tibble: 4 × 46
#> date patient age gender ward mo PEN OXA FLC AMX
#> <date> <chr> <dbl> <chr> <chr> <mo> <rsi> <rsi> <rsi> <rsi>
#> 1 2002-08-31 149442 80 F ICU B_STPHY_AURS R NA S R
#> 2 2002-09-24 CF9318 29 M ICU B_CMPYL_JEJN NA NA NA NA
#> 3 2002-10-11 871360 78 M Clinical B_STPHY_EPDR R NA S NA
#> 4 2002-08-28 390178 57 M Clinical B_STRPT_SLVR S NA NA S
#> # … 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 403631 2007-07-31 C FALSE
#> 2 304347 2002-11-04 C FALSE
#> 3 964129 2016-06-21 A FALSE
#> 4 189363 2004-03-17 C FALSE
#> 5 80C025 2007-11-13 A FALSE
#> 6 284FFF 2010-03-31 A FALSE
#> 7 A76045 2015-10-06 C FALSE
#> 8 959835 2013-11-16 B FALSE
#> 9 277241 2005-09-01 C FALSE
#> 10 696587 2004-11-16 C 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 [179]
#> ward date patient new_index new_logical
#> <chr> <date> <chr> <dbl> <lgl>
#> 1 Clinical 2007-07-31 403631 1 TRUE
#> 2 Clinical 2002-11-04 304347 1 TRUE
#> 3 Clinical 2016-06-21 964129 1 TRUE
#> 4 Clinical 2004-03-17 189363 1 TRUE
#> 5 Clinical 2007-11-13 80C025 1 TRUE
#> 6 Clinical 2010-03-31 284FFF 1 TRUE
#> 7 ICU 2015-10-06 A76045 1 TRUE
#> 8 Clinical 2013-11-16 959835 1 TRUE
#> 9 ICU 2005-09-01 277241 1 TRUE
#> 10 ICU 2004-11-16 696587 1 FALSE
#> # … 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 112 14 57 81
#> 2 ICU 53 13 37 44
#> 3 Outpatient 14 8 11 12
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 [186]
#> patient mo ward flag_episode
#> <chr> <mo> <chr> <lgl>
#> 1 403631 B_ESCHR_COLI Clinical TRUE
#> 2 304347 B_STRPT_PNMN Clinical TRUE
#> 3 964129 B_SERRT_MRCS Clinical TRUE
#> 4 189363 B_STPHY_AURS Clinical TRUE
#> 5 80C025 B_ESCHR_COLI Clinical TRUE
#> 6 284FFF B_STPHY_EPDR Clinical TRUE
#> 7 A76045 B_STPHY_EPDR ICU TRUE
#> 8 959835 B_STPHY_EPDR Clinical TRUE
#> 9 277241 B_STPHY_AURS ICU TRUE
#> 10 696587 B_ESCHR_COLI ICU FALSE
#> # … with 190 more rows
# }