4
Peña ME. El tamaño muestral en la invesꢁgación quirúrgica. Rev Argent Cir. 2026;118(2):e-cymmep
■
ENGLISH VERSION
Sample size calculaꢀon is an essenꢀal
What types of research quesꢀons exist?
component of the methodological design in medical
research. In scienꢀfic studies, reporꢀng the sample
size improves the interpretability of the results,
strengthening both the internal validity and ethical
legiꢀmacy of the study. This aspect is parꢀcularly
important in surgery, given that many clinical events
are relaꢀvely rare and surgical techniques can vary
considerably from surgeon to surgeon.
Generally speaking, research quesꢀons can
be classified as descripꢀve and analyꢀcal. Each type of
quesꢀon requires a specific methodological approach
anddifferentstrategiesforcalculaꢀngthesamplesize3
–5
.
Descripꢀve quesꢀons
Despite its significance, failure to calculate the
The aim of descripꢀve quesꢀons is to
characterize or describe how a phenomenon manifests
itself or how a variable is distributed within a given
populaꢀon, without aꢄempꢀng to establish causal
relaꢀonships. Case series or cross-secꢀonal studies are
commonly used for this type of quesꢀon.
sample size remains a common shortcoming in surgical
1
,2
publicaꢀons . The purpose of this arꢀcle is to review
the definiꢀon, significance, indicaꢀons, and general
principles of sample size calculaꢀon.
What is sample size calculaꢀon?
Examples:
Sample size calculaꢀon involves esꢀmaꢀng
the number of parꢀcipants or observaꢀon units that
must be included in a study to adequately answer the
▪ What is the average length of hospital stay following a
laparoscopic left hemicolectomy?
▪ What is the mortality rate among patients diagnosed
with severe acute pancreatitis?
2
,4
research quesꢀon posed (see below) .
In surgery, the units of observaꢀon are usually
paꢀents or surgical procedures, as well as hospitals,
surgeons, or specific clinical episodes. For example,
a study analyzing all appendectomies performed in
In these cases, the sample size is typically
calculated to esꢀmate a populaꢀon parameter (mean
or proporꢀon) with a specified level of precision.
2
025 at a community-based hospital included 120
procedures. In this case, the sample size consists of 120
appendectomies (n=120).
Analyꢀcal quesꢀons
Analyꢀcal quesꢀons are designed to evaluate
the relaꢀonship between two or more variables to
determine whether there is an associaꢀon, a difference,
or a possible causal effect between them.
In this type of study, a hypothesis is formulated
and expressed as follows:
Why is it important to calculate the sample size?
When the sample size is not preliminarily
calculated, the number of observaꢀon units included
may be greater or fewer than necessary. When more
parꢀcipants than necessary are included, costs,
resource usage, and research ꢀme are unnecessarily
increased. Conversely, if the sample size is too small,
the study may not have sufficient power to detect a
real difference between the groups being analyzed.
▪ Null hypothesis (H0): There is no difference or associa-
tion between the variables.
▪ Alternative hypothesis (H1): There is a difference or
association.
This situaꢀon is known as a type II error or lack of
The study design aims to test the null
hypothesis using staꢀsꢀcal analysis. If the results
reject the null hypothesis, the alternaꢀve hypothesis
is accepted, suggesꢀng the existence of a staꢀsꢀcally
significant difference or associaꢀon. These types of
quesꢀons are typically addressed using analyꢀcal
designs such as prospecꢀve or retrospecꢀve cohort
studies, case-control studies, or clinical trials.
staꢀsꢀcal power3
–5
.
Moreover, an ethical dimension is also
implicated. In research studies, paꢀents provide
sensiꢀve informaꢀon with the aim of contribuꢀng to
scienꢀfic knowledge. The design of studies in which
the probability of answering the research quesꢀon is
very low may result in an inappropriate use of the data.
For instance, a retrospecꢀve analysis of an extremely
rare condiꢀon, such as stump appendiciꢀs, based on
only three available cases, would not likely yield robust
conclusions.
For example, what is the surgical site
infecꢀon rate for open appendectomies compared to
laparoscopic appendectomies? In this case, the null
hypothesis would state that there are no differences