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$date, episode_days = 60) # indices
#> [1] 20 35 5 21 40 9 3 60 33 52 45 42 58 43 38 1 10 12 9 37 3 42 60 7 60
#> [26] 56 59 49 50 32 50 60 58 29 18 57 20 43 52 29 24 45 29 54 37 25 58 42 12 40
#> [51] 59 25 5 52 59 25 59 48 7 48 57 14 59 28 55 49 12 24 6 7 11 38 48 7 37
#> [76] 45 24 56 54 10 47 29 55 7 1 11 35 9 42 16 58 23 2 12 38 3 33 14 17 38
#> [101] 52 22 34 26 10 5 51 6 31 41 30 4 11 25 27 24 54 29 12 32 23 49 20 10 18
#> [126] 13 51 54 31 7 35 50 44 38 16 5 39 30 15 46 3 60 19 3 40 15 45 46 43 8
#> [151] 9 43 39 5 9 18 33 42 55 55 42 56 17 17 34 55 36 26 26 15 1 46 26 53 40
#> [176] 19 59 14 34 25 60 47 5 29 15 19 23 54 28 7 47 9 39 23 22 9 52 24 61 21
is_new_episode(df$date, episode_days = 60) # TRUE/FALSE
#> [1] TRUE FALSE FALSE TRUE FALSE FALSE FALSE FALSE TRUE FALSE FALSE TRUE
#> [13] TRUE FALSE FALSE FALSE TRUE FALSE FALSE FALSE FALSE FALSE TRUE FALSE
#> [25] FALSE TRUE FALSE FALSE TRUE TRUE FALSE FALSE FALSE TRUE FALSE FALSE
#> [37] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE TRUE TRUE FALSE FALSE
#> [49] FALSE TRUE FALSE FALSE TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
#> [61] TRUE FALSE TRUE TRUE FALSE FALSE TRUE TRUE FALSE FALSE TRUE TRUE
#> [73] TRUE TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
#> [85] TRUE FALSE FALSE TRUE FALSE FALSE FALSE FALSE TRUE FALSE FALSE FALSE
#> [97] FALSE FALSE TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE TRUE
#> [109] TRUE TRUE TRUE TRUE FALSE FALSE TRUE FALSE TRUE FALSE FALSE FALSE
#> [121] FALSE TRUE FALSE FALSE FALSE TRUE TRUE FALSE FALSE FALSE TRUE FALSE
#> [133] TRUE FALSE TRUE FALSE FALSE FALSE FALSE TRUE FALSE FALSE FALSE TRUE
#> [145] FALSE TRUE TRUE FALSE TRUE TRUE FALSE FALSE TRUE FALSE FALSE TRUE
#> [157] FALSE FALSE TRUE FALSE FALSE FALSE FALSE FALSE TRUE FALSE TRUE FALSE
#> [169] TRUE FALSE FALSE FALSE FALSE TRUE FALSE TRUE FALSE TRUE FALSE FALSE
#> [181] FALSE TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
#> [193] FALSE TRUE TRUE FALSE TRUE FALSE TRUE FALSE
# filter on results from the third 60-day episode only, using base R
df[which(get_episode(df$date, 60) == 3), ]
#> # A tibble: 5 × 46
#> date patient age gender ward mo PEN OXA FLC AMX
#> <date> <chr> <dbl> <chr> <chr> <mo> <rsi> <rsi> <rsi> <rsi>
#> 1 2002-11-12 CF9318 29 M ICU B_STPHY_CONS R NA R R
#> 2 2002-11-11 D80753 74 F Outpatie… B_STPHY_CONS R NA S NA
#> 3 2002-11-28 600057 88 M Outpatie… B_STPHY_AURS R NA S R
#> 4 2002-12-13 285137 78 F ICU B_ESCHR_COLI R NA NA NA
#> 5 2002-11-04 304347 62 M Clinical B_STRPT_PNMN 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 561303 2006-08-11 B TRUE
#> 2 F54287 2010-07-05 C FALSE
#> 3 6BC362 2003-04-21 C TRUE
#> 4 557456 2006-10-24 B FALSE
#> 5 F56314 2011-10-17 B FALSE
#> 6 406502 2004-02-29 B FALSE
#> 7 CF9318 2002-11-12 C FALSE
#> 8 8DD375 2017-09-12 C FALSE
#> 9 650870 2009-11-12 C TRUE
#> 10 0DBF93 2015-10-12 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 [180]
#> ward date patient new_index new_logical
#> <chr> <date> <chr> <dbl> <lgl>
#> 1 Clinical 2006-08-11 561303 1 TRUE
#> 2 Clinical 2010-07-05 F54287 1 TRUE
#> 3 ICU 2003-04-21 6BC362 1 FALSE
#> 4 ICU 2006-10-24 557456 1 TRUE
#> 5 Outpatient 2011-10-17 F56314 1 TRUE
#> 6 Clinical 2004-02-29 406502 1 TRUE
#> 7 ICU 2002-11-12 CF9318 1 TRUE
#> 8 ICU 2017-09-12 8DD375 1 TRUE
#> 9 Outpatient 2009-11-12 650870 1 TRUE
#> 10 Clinical 2015-10-12 0DBF93 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 72
#> 2 ICU 60 10 36 42
#> 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 [189]
#> patient mo ward flag_episode
#> <chr> <mo> <chr> <lgl>
#> 1 561303 B_STPHY_CONS Clinical TRUE
#> 2 F54287 B_STRPT_ANGN Clinical TRUE
#> 3 6BC362 B_ENTRC ICU TRUE
#> 4 557456 B_BCTRD_FRGL ICU TRUE
#> 5 F56314 B_ESCHR_COLI Outpatient TRUE
#> 6 406502 B_STRPT_PNMN Clinical TRUE
#> 7 CF9318 B_STPHY_CONS ICU TRUE
#> 8 8DD375 B_ESCHR_COLI ICU TRUE
#> 9 650870 B_ESCHR_COLI Outpatient TRUE
#> 10 0DBF93 B_STPHY_EPDR Clinical TRUE
#> # … with 190 more rows
# }