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$date, episode_days = 60) # indices
#> [1] 26 38 48 5 5 4 36 61 56 6 56 45 10 48 13 11 49 55 54 24 4 25 44 62 54
#> [26] 55 16 26 28 11 4 32 57 7 55 21 56 49 16 6 6 35 1 55 50 26 54 17 39 8
#> [51] 36 6 41 59 47 18 8 47 13 24 18 24 3 59 49 24 32 34 59 20 60 2 25 17 4
#> [76] 12 36 48 29 4 22 2 19 34 61 33 15 44 9 60 7 32 20 27 16 45 42 42 56 19
#> [101] 60 43 3 36 13 30 8 16 12 22 27 14 22 36 25 4 8 29 14 45 43 12 20 52 3
#> [126] 61 30 19 29 28 48 21 57 36 52 47 40 12 37 37 51 53 50 24 57 43 60 45 51 19
#> [151] 19 12 26 43 23 50 27 63 40 57 55 7 8 46 32 20 31 22 57 26 45 61 25 29 39
#> [176] 9 3 31 38 5 22 26 42 31 18 10 58 33 51 41 31 39 31 61 30 25 32 44 35 6
is_new_episode(df$date, episode_days = 60) # TRUE/FALSE
#> [1] FALSE TRUE FALSE FALSE TRUE FALSE FALSE FALSE TRUE FALSE FALSE FALSE
#> [13] FALSE TRUE FALSE FALSE TRUE FALSE TRUE FALSE FALSE FALSE FALSE TRUE
#> [25] FALSE TRUE TRUE FALSE FALSE TRUE FALSE FALSE FALSE FALSE FALSE TRUE
#> [37] FALSE FALSE FALSE FALSE TRUE TRUE TRUE FALSE FALSE FALSE FALSE FALSE
#> [49] FALSE TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE TRUE TRUE TRUE
#> [61] FALSE FALSE TRUE TRUE FALSE FALSE FALSE TRUE FALSE TRUE FALSE TRUE
#> [73] FALSE TRUE TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
#> [85] FALSE FALSE TRUE TRUE FALSE FALSE FALSE FALSE FALSE TRUE FALSE FALSE
#> [97] FALSE TRUE FALSE TRUE FALSE TRUE FALSE FALSE FALSE FALSE FALSE FALSE
#> [109] FALSE TRUE FALSE TRUE FALSE TRUE FALSE FALSE FALSE TRUE FALSE FALSE
#> [121] FALSE FALSE FALSE TRUE FALSE TRUE TRUE FALSE FALSE TRUE FALSE FALSE
#> [133] TRUE FALSE FALSE FALSE TRUE TRUE TRUE FALSE TRUE TRUE TRUE FALSE
#> [145] FALSE FALSE TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE TRUE FALSE
#> [157] FALSE TRUE FALSE FALSE FALSE TRUE FALSE TRUE FALSE FALSE TRUE FALSE
#> [169] FALSE TRUE TRUE FALSE FALSE FALSE FALSE TRUE FALSE FALSE FALSE FALSE
#> [181] FALSE FALSE FALSE FALSE TRUE TRUE TRUE TRUE FALSE TRUE FALSE TRUE
#> [193] FALSE FALSE FALSE TRUE TRUE 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: 4 × 46
#> date patient age gender ward mo PEN OXA FLC AMX
#> <date> <chr> <dbl> <chr> <chr> <mo> <rsi> <rsi> <rsi> <rsi>
#> 1 2002-12-13 285137 78 F ICU B_ESCHR_COLI R NA NA NA
#> 2 2003-01-27 F35553 51 M ICU B_STPHY_EPDR R NA S NA
#> 3 2003-01-25 088256 73 F ICU B_STPHY_CONS R NA R R
#> 4 2003-01-08 783073 83 M Clinical B_STPHY_CONS R NA S 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 A42180 2008-04-21 C FALSE
#> 2 116866 2011-11-07 B TRUE
#> 3 786327 2014-06-02 B FALSE
#> 4 F71508 2003-08-14 A FALSE
#> 5 900485 2003-07-17 B FALSE
#> 6 4DD722 2003-06-02 B FALSE
#> 7 E31724 2011-07-05 B FALSE
#> 8 650215 2017-05-27 B FALSE
#> 9 F61180 2016-02-15 C FALSE
#> 10 1B0933 2003-09-28 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 [189]
#> ward date patient new_index new_logical
#> <chr> <date> <chr> <dbl> <lgl>
#> 1 Clinical 2008-04-21 A42180 1 TRUE
#> 2 Clinical 2011-11-07 116866 1 TRUE
#> 3 Outpatient 2014-06-02 786327 1 TRUE
#> 4 Clinical 2003-08-14 F71508 1 TRUE
#> 5 Clinical 2003-07-17 900485 1 TRUE
#> 6 ICU 2003-06-02 4DD722 1 TRUE
#> 7 ICU 2011-07-05 E31724 1 TRUE
#> 8 ICU 2017-05-27 650215 1 TRUE
#> 9 ICU 2016-02-15 F61180 1 TRUE
#> 10 Clinical 2003-09-28 1B0933 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 112 15 52 73
#> 2 ICU 65 13 36 50
#> 3 Outpatient 12 7 10 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] TRUE
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 [197]
#> patient mo ward flag_episode
#> <chr> <mo> <chr> <lgl>
#> 1 A42180 B_ENTRC_AVIM Clinical TRUE
#> 2 116866 B_STPHY_CONS Clinical TRUE
#> 3 786327 B_STRPT_EQUI Outpatient TRUE
#> 4 F71508 B_STRPT_GRPB Clinical TRUE
#> 5 900485 B_STPHY_CONS Clinical TRUE
#> 6 4DD722 B_ESCHR_COLI ICU TRUE
#> 7 E31724 B_STPHY_CONS ICU TRUE
#> 8 650215 B_STRPT_INFN ICU TRUE
#> 9 F61180 B_ESCHR_COLI ICU TRUE
#> 10 1B0933 B_ENTRC Clinical TRUE
#> # … with 190 more rows
# }