RIPASA Score For Diagnosis of AA

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

Original Article Singapore Med J 2010; 51(3) : 220

Development of the RIPASA score: a


new appendicitis scoring system for the
diagnosis of acute appendicitis
Chong C F, Adi M I W, Thien A, Suyoi A, Mackie A J, Tin A S, Tripathi S, Jaman N H,
Tan K K, Kok K Y, Mathew V V, Paw O, Chua H B, Yapp S K

Department of
Surgery, ABSTRACT Conclusion: The new appendicitis scoring system
Raja Isteri Pengiran
Anak Saleha Introduction: Acute appendicitis is one of the most looked promising when applied to our settings,
Hospital,
Bandar Seri common surgical emergencies. The Alvarado and and had a better sensitivity and specificity
Begawan BA1710,
Brunei Darussalam
modified Alvarado scores have been developed to than the Alvarado score when applied to Asian
aid diagnosis, but both scoring systems have poor populations. A significant reduction in the
Chong CF, FRCS,
FRCSE, MD sensitivity and specificity when applied in Middle negative appendicectomy rate was also predicted.
Specialist Surgeon
Eastern and Asian populations. The aim of this A prospective evaluation of this new appendicitis
Thien A, MRCS study was to develop a new scoring system that is scoring system, referred to as the RIPASA score,
Senior Medical
Officer suitable for the local population. is ongoing.
Suyoi A, MRCS
Senior Medical
Officer
Methods: Clinical data from 312 patients who had Keywords: acute appendicitis, appendicectomy,
undergone an emergency appendicectomy was diagnostic techniques, surgical, symptoms
Mackie AJ, FRCS
Senior Medical retrospectively collected and used to generate Singapore Med J 2010; 51(3): 220-225
Officer
15 parameters. The probability was calculated
Tin AS, MRCS and a score of 0.5, 1.0 or 2.0 was allocated to each INTRODUCTION
Senior Medical
Officer parameter. The receiver operating curve (ROC), Acute appendicitis is one of the most common
Tripathi S, MS sensitivity, specificity, positive predictive value surgical emergencies, with a lifetime prevalence rate
Medical Officer
(PPV) and negative predictive value (NPV) of of approximately one in seven.(1) The incidence is
Jaman NH, MBChB the new scoring system were derived using the 1.5–1.9 per 1,000 in the male and female population,
Medical Officer
StatsDirect statistical software. and is approximately 1.4 times greater in men than in
Tan KK, FRCS
Specialist Surgeon women.(2)
Results : The 15 parameters and the scores The diagnosis of acute appendicitis is based
Kok KY, FRCS,
FAMS generated were age (less than 40 years is 1 point; purely on clinical history and examination combined
Specialist Surgeon
greater than 40 years is 0.5 point), gender (male is with laboratory investigations such as elevated white
Mathew VV, FRCS
Specialist Surgeon
1 point; female is 0.5 point), right iliac fossa (RIF) cell count. Despite being a common problem, acute
pain (0.5 point), migration of pain to RIF (0.5 appendicitis remains a difficult diagnosis to establish,
Paw O, FRCS
Specialist Surgeon point), nausea and vomiting (1 point), anorexia particularly among the young, the elderly and females
Chua HB, FRCS (1 point), duration of symptoms (less than 48 of reproductive age, where a host of other genitourinary
Specialist Surgeon hours is 1 point; more than 48 hours is 0.5 point), and gynaecological inflammatory conditions can present
Yapp SK, FRCS RIF tenderness (1 point), guarding (2 points), with signs and symptoms that are similar to those of acute
Specialist Surgeon
rebound tenderness (1 point), Rovsing’s sign (2 appendicitis.(3) A delay in performing an appendicectomy
Institute of Medicine,
University Brunei
points), fever (1 point), raised white cell count (1 in order to improve its diagnostic accuracy increases
Darussalam, point), negative urinalysis (1 point) and foreign the risk of appendicular perforation and sepsis, which
Jalan Tungku Link,
Gadong BE1410, national registration identity card (1 point). The in turn increases morbidity and mortality.(4) The
Brunei Darussalam
optimal cut-off threshold score from the ROC was opposite is also true, where with reduced diagnostic
Adi MIW 7.5, with a sensitivity of 88 percent, a specificity accuracy, the negative or unnecessary appendicectomy
Medical Student
of 67 percent, a PPV of 93 percent and an NPV of rate is increased, and this is generally reported to be
Correspondence to:
Mr Chong Chee Fui 53 percent. The negative appendicectomy rate approximately 20%–40%.(5)
Tel: (673) 224 2424
ext 280
decreased significantly from 16.3 percent to 6.9 Diagnostic accuracy can be further improved through
Fax: (673) 233 3270 percent, which was a 9.4 percent reduction (p is the use of ultrasonography or computed tomography
Email: chong_chee_
[email protected] 0.0007). imaging.(6) However, these modalities are costly and may
Singapore Med J 2010; 51(3) : 221

Table I. Patient demographics.


Positive Negative
Demographic No. (%) (n = 312) Appendicectomy Appendicectomy

Male: female ratio 180:132 120 42.9%


41.4%
Mean age ± SD (years) 26 ± 13.5
100
Positive histology for acute appendicitis 261 (83.7)

No. of patients
Negative histology for acute appendicitis 51 (16.3) 80
Negative appendicectomy rate (%) 16.3
Laparoscopic appendicectomy 42 (13.5) 60
Mean hospital stay ± SD (days) 4.6 ± 3.8 14.9%
Postoperative complications (%) 22 (7.1) 40
Superficial wound infection 13 (4.2)
20
Wound haematoma 2 (0.6) 0.8%
Wound pain 4 (1.3) 0
Intra-abdominal sepsis 3 (1.0) 0–20 20–39.9 40–60 > 60
No. of patients discharged alive 312 (100) Age (years)

SD: standard deviation Fig. 1 Distribution of patients who underwent an emergency


appendicectomy according to age.

not be easily available when they are required. Making


arrangements for these diagnostic modalities may lead to
< 48 hrs > 48 hrs
further delays in diagnosis and surgery. Several scoring
systems have been developed to aid in the diagnosis of 200
180
acute appendicitis. The Alvarado score and the modified
160
Alvarado score are the two most commonly used scoring 140
No. of patients

systems.(5,7) The reported sensitivity and specificity 120 67%


for the Alvarado and the modified Alvarado scores 100
80
range from 53%–88% and 75%–80%, respectively.(5,7)
60
33%
However, these scoring systems were developed in 40
western countries, and several studies have reported 20
55% 45%
very low sensitivity and specificity when these scores 0
Positive Appendicectomy Negative Appendicectomy
are applied to a population with a completely different
Fig. 2 Distribution of patients who underwent an emergency
ethnic origin and diet.(8,9) Thus, the objective of this appendicectomy according to the duration of symptoms.
study was to develop an appendicitis scoring system that
is more applicable to the Southeast Asian region.
records, which were collected from the Medical Record
METHODS Department of RIPAS Hospital. The medical records of
This was a retrospective study consisting of 400 patients the other 77 patients were not traceable, and hence, these
who had undergone an appendicectomy between patients were excluded from the study. Out of the 323
October 2006 and May 2008, and who were identified patients, only 312 patients satisfied our inclusion and
from the operation note database of the Department exclusion criteria. The other 11 patients were excluded
of Surgery, Raja Isteri Pengiran Anak Saleha (RIPAS) as they presented with non-RIF pain.
Hospital, Brunei Darussalem. The inclusion criteria The data collected included the patients’
were patients of all age groups who presented with right demographics (national registration identity card [NRIC]
iliac fossa (RIF) pain suspected to be acute appendicitis, number, age and gender), the presenting symptoms
and who had undergone emergency appendicectomy (RIF pain, the migration of pain to the RIF, nausea and
as the primary procedure. Patients presenting with any vomiting, anorexia and the duration of symptoms),
form of non-RIF pain, such as lower abdominal pain or clinical signs (RIF tenderness, guarding, rebound
right upper quadrant pain, and those who had undergone tenderness, Rovsing’s sign and fever) and laboratory
other emergency laparotomy where appendicectomy investigations (elevated white cell count and negative
was also performed as part of the procedure, or elective urinalysis). The inclusion of these 15 parameters was
appendicectomy, were excluded. Ethical approval for the agreed upon by a panel of general surgeons at RIPAS
study was obtained from the Ethics Committee Review Hospital. These 15 parameters form the basis of the new
Board of RIPAS Hospital. appendicitis scoring system. The probability of each
Of the 400 patients, only 323 had complete medical parameter was calculated and scores of 0.5, 1.0 or 2.0
Singapore Med J 2010; 51(3) : 222

Table II. The probability of acute appendicitis for each parameter, with the scoring of parameters based on
probabilities and extra weightage.
Scoring Elements Probability Odds ratio Score Missing data (%)

Male 0.90 3.10 1.0 0.0


Female 0.75 - 0.5 -
Age < 39.9 yrs 0.83 0.85 1.0 0.0
Age > 40 yrs 0.85 - 0.5 -
RIF pain 0.70 - 0.5 0.0
Migration of RLQ pain 0.83 1.03 0.5 18.0
Anorexia 0.90 0.50 1.0 54.0
Nausea & Vomiting 0.90 0.29 1.0 1.0
Duration of symptoms < 48 hrs 0.86 0.60 1.0 0.0
Duration of symptoms > 48 hrs 0.79 - 0.5 0.0
RIF tenderness 0.84 1.18 1.0 0.3
RIF guarding† 0.92 0.21 2.0 7.0
Rebound tenderness 0.88 0.59 1.0 36.0
Rovsing’s Sign† 0.91 0.47 2.0 84.0
Fever 0.94 0.22 1.0 2.0
Raised WCC 0.86 0.42 1.0 0.0
Negative urinalysis* 0.87 0.54 1.0 13.0
Foreign NRIC** 0.96 5.75 1.0 0.0
Minimum Total Score - - 2 -
Maximum Total Score - - 16 -


Extra weightage provided by agreement of a panel of general surgeons.
* Negative urinalysis: absence of blood, neutrophils or bacteria.
** Additional parameter.
RIF: right iliac fossa; RLQ: right lower quadrant; WCC: white cell count; NRIC: national registration identity card

points were allocated to each parameter based on its had undergone emergency appendicectomy, as shown
probability, with extra weightage provided to two clinical in Table I. The mean age of the group was 26.0 ± 13.5
signs: guarding and Rovsing’s signs. Confirmation of years, with a male to female ratio of 180:132 (1.4:1). A
acute appendicitis as the final diagnosis was obtained positive diagnosis of acute appendicitis was confirmed
from a histological analysis of the resected appendix at on histological analysis of the resected appendix in
the Department of Histopathology at RIPAS Hospital. 261 patients, while 51 patients had a normal appendix,
The binomial data was analysed using a non- indicating a negative appendicectomy rate of 16.3%.
parametric chi-square test. The probability and odds The mean duration of hospital stay was 4.6 ± 3.8 days.
ratio for each parameter were derived using logistic The rate of postoperative complications was 7%, and
regression analysis. The receiver operating curve consisted mainly of superficial wound infections, as
(ROC) at the optimal cut-off threshold score for the shown in Table I. All 312 patients were discharged
new appendicitis scoring system was derived using the alive.
StatsDirect statistical software version 2.7.2 (StatsDirect 84.3% of the patients with acute appendicitis
Ltd, Cheshire, UK). The sensitivity, specificity, positive were < 40 years of age, while 15.7% were > 40 years
predictive value (PPV) and negative predictive value of age (Fig. 1). Hence, for the development of the new
(NPV) at the optimal cut-off threshold score were appendicitis scoring system, age was divided into two
also derived from the ROC.(10) The predicted negative groups: < 40 years and > 40 years of age. Similarly, the
appendicectomy rates for the new appendicitis scoring majority of patients with acute appendicitis presented
system were also derived and compared with the negative within 48 hours of appearance of symptoms (Fig. 2),
appendicectomy rate from the raw data. The intra- and the duration of symptoms in the new appendicitis
observer and inter-observer variability of the dataset scoring system was divided into two groups: < 48 hours
collected were assessed using correlation and regression and > 48 hours.
analysis as well as Bland-Altman plots in ten randomly The parameters included in the new appendicitis
selected patients.(11) scoring system consisted of age, gender, RIF pain, the
migration of pain to the RIF, nausea and vomiting,
RESULTS anorexia, the duration of symptoms, RIF tenderness,
The study population consisted of 312 patients who guarding, rebound tenderness, Rovsing’s sign, fever,
Singapore Med J 2010; 51(3) : 223

Sensitivity the ROC analysis was 7.5, as shown in Fig. 3. Based on


1.00 this optimal cut-off threshold, the calculated sensitivity
and specificity were 88.46% (95% confidence interval
0.75 [CI] 83.94%–92.08%) and 66.67% (95% CI 52.08%–
Optimal cut-off 79.24%), respectively (Fig. 3). The PPV and NPV were
threshold total
score = 7.5 93.00% and 53.00%, respectively (Fig. 3). The diagnostic
0.50 accuracy was 80.50% (95% CI 73.35%–87.65%) (Fig.
3). The predicted negative appendicectomy rate at the
0.25 optimal cut-off threshold score of 7.5 was 6.9%, which
was a 9.3% reduction from the raw data (16.3%), and
this was statistically significant (p = 0.0007).
0.00 The correlation regression coefficients for the intra-
0.00 0.25 0.50 0.75 1.00
1-Specificity observer and inter-observer variability analysis of the
Fig. 3 ROC plot for the new appendicitis scoring system. The dataset were 0.93 and 0.88, respectively (Figs. 4a & 5a).
optimal cut-off threshold score is 7.5, with a sensitivity and Figs. 4b and 5b show the Bland-Altman plots for both
specificity of 0.88 and 0.67, respectively, and a diagnostic accuracy
intra-observer and inter-observer variability, showing
of 0.81. The positive predictive value and negative predictive
value are 0.93 and 0.53, respectively. that the majority of the data were within ± 1 standard
deviation (SD) of the average difference. Both the
correlation regression analysis and Bland-Altman plots
elevated white cell count, negative urinalysis and a indicate that the dataset was reliable.
foreign NRIC included as an additional parameter
because of the high probability of acute appendicitis seen DISCUSSION
in foreign nationals presenting with RIF pain (Table II). Acute appendicitis is one of the most commonly
The probabilities for acute appendicitis were calculated encountered surgical emergencies, especially by junior
for each of the 15 parameters, as shown in Table II. doctors on call, with emergency appendicectomy making
Scoring of the parameters was done based on the up 10% of all emergency abdominal surgeries.(12,13)
probability of acute appendicitis. Male gender was Several scoring systems, such as the Alvarado and
found to have a higher probability than female gender, modified Alvarado scoring system, have been introduced
and hence was scored with 1.0 point while female since 1986 to help with the clinical decision-making
gender was given a score of 0.5 point. As more than process in achieving an accurate diagnosis of acute
84% of patients with acute appendicitis were < 40 years appendicitis in the fastest and cheapest way.(5,7) However,
of age (Fig. 1), despite the slightly lower probability these two scoring systems were created in the West, and
compared with an age > 40 years, having an age < 40 when applied in different environments, such as the
years was scored with 1.0 point, while an age > 40 years Middle East and Asia, the sensitivity and specificity
was scored with 0.5 point. Both the presence of RIF levels achieved were very low.(8,9) Khan et al applied
pain and the migration of pain to RIF were combined the Alvarado scoring system in an Asian population and
for a score of 1.0 point; thus, a score of 0.5 point was only achieved a sensitivity and specificity of 59% and
allocated to each of these parameters. A duration of 23%, respectively, with a negative appendicectomy rate
symptoms of < 48 hours showed a higher probability of of 15.6%.(9) Another study by Al-Hashemy et al in 2004
acute appendicitis, and was scored with 1.0 point, while using the modified Alvarado scoring system in a Middle
a duration of symptoms > 48 hours was scored with 0.5 Eastern population reported a similarly low sensitivity
point. Both signs of localised guarding and Rovsing’s of 53.8% and a specificity of 80%.(8) The sensitivity of
sign were weighted highly by our panel of local general the Alvarado score achieved when applied in an oriental
surgeons, as the presence of these two clinical signs was population, at the suggested cut-off threshold of 7.0, was
highly indicative of acute appendicitis. Hence, these similarly low at 50.6%, but achieved a high specificity
two parameters were scored with 2.0 points each. The of 94.5%.(14) However, this improved when the cut-off
remaining parameters (nausea and vomiting, anorexia, threshold was lowered to 6.0, with a sensitivity and
RIF tenderness, rebound tenderness, fever, elevated specificity of 88.3% and 94.5%, respectively, suggesting
white cell count, negative urinalysis and foreign NRIC) a definite ethnic difference with regard to the Alvarado
were all scored with 1.0 point each (see Table II). score.(14)
The optimal cut-off threshold score derived from Both the Alvarado and modified Alvarado scores
Singapore Med J 2010; 51(3) : 224

16 16 y = 1.0475x

14 14

12 y = 0.970x 12

10 10
Total score 1

Total score 1
8 8

6 6

4 4

2 2
Correlation coefficient, r = 0.927 Correlation coefficient, r = 0.876
0 0
0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16
Total score 2 Total score 2
Fig. 4a The correlation regression analysis of intra-observer Fig. 5a The correlation regression analysis of inter-observer
variability. variability.

2.5 1
2
0.5 +1 SD
1.5
+1 SD 0
Score difference

1
Score difference

-0.5 Mean
0.5 Mean
0 -1
-0.5 -1 SD
-1.5 -1 SD

-1 -2
-1.5 -2.5
-2 -3
-2.5 -3.5
0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16
Mean score, (TS1 + TS2)/2 Mean score, (TS1 + TS2)/2
Fig. 4b The Bland-Altman plot of intra-observer variability. Fig. 5b The Bland-Altman plot of inter-observer variability.

lack parameters that have been shown to be important are earlier indicators of a local inflammatory process
determinants in the diagnosis of acute appendicitis, such such as acute appendicitis, while rebound tenderness
as age, gender and the duration of symptoms. Wani et is a much later sign when the peritoneum is involved
al have shown that the sensitivity and specificity of the with peritonism. Negative urinalysis was also included
Alvarado scoring system vary with age, gender and the to exclude urinary causes of RIF pain, as 60% of our
duration of symptoms.(15) Our study has confirmed the general surgical admission was urological in nature.
presence of age differences (Fig. 1) and differences Lastly, foreign NRIC was included as an additional
in the duration of symptoms (Fig. 2) in histologically parameter as the authors had found a high probability
confirmed cases of acute appendicitis. Furthermore, (0.8) of acute appendicitis in foreign nationals presenting
gender differences in the occurrence of acute appendicitis with RIF pain. There is a large foreign labour workforce
were also found in our study, with male patients in Brunei Darussalam who must pay for their medical
being 1.4 times more likely than female patients to be treatment at RIPAS Hospital. For this reason, foreign
diagnosed, and this is in keeping with published data.(2) nationals tend to present much later when the symptoms
This new appendicitis scoring system includes the three are more severe.
parameters mentioned above as well as four other new The minimum and maximum total scores achievable
parameters deemed important in our local settings, with this new appendicitis scoring system were 2
including clinical signs of RIF guarding, Rovsing’s sign, and 16, respectively. The sensitivity and specificity
negative urinalysis and foreign NRIC status. achieved were 88% and 67%, respectively, with a
Guarding and Rovsing’s sign were included as the diagnostic accuracy of 81%, which is comparable to
panel of general surgeons felt that these two clinical signs the Alvarado score when the latter was applied in a
Singapore Med J 2010; 51(3) : 225

Western population.(7) This was a definite improvement is ongoing, and the authors aimed to recruit 100–150
from the Alvarado score (sensitivity 50.6%–59.0%, patients prospectively.
specificity 23.0%–94.5%) and modified Alvarado score
(sensitivity 53.8%, specificity 80%) when applied to REFERENCES
1. Stephens PL, Mazzucco JJ. Comparison of ultrasound and the
Middle Eastern, Asian or Oriental populations.(8,9,14) The
Alvarado score for the diagnosis of acute appendicitis. Conn Med
PPV and NPV for the new appendicitis score, at 93% 1999; 63:137-40.
and 53%, respectively, are also comparable to those 2. Cuscheri A. The small intestine and vermiform appendix. In:
achieved with the Alvarado and modified Alvarado Cuschieri A, Giles GR, Mossa AR, eds. Essential Surgical Practice.
3rd ed. Oxford: Butterworth-Heinermann, 1995: 1297-329.
scores.(8,9,14) Using the new appendicitis scoring system, 3. Gilmore OJ, Browett JP, Griffin PH, et al. Appendicitis and
the predicted negative appendicectomy rate was 6.9%, mimicking conditions. A prospective study. Lancet 1975;
which was a 9.4% reduction from the raw data, and 2:421-4.
4. Velanovich V, Satava R. Balancing the normal appendectomy
highly significant statistically (p = 0.0007).
rate with the perforated appendicitis rate: implications for quality
This new appendicitis scoring system was assurance. Am Surg 1992; 58:264-9.
specifically developed for our local patient group, but it 5. Kalan M, Talbot D, Cunliffe WJ, Rich AJ. Evaluation of the
modified Alvarado score in the diagnosis of acute appendicitis: a
is likely to be applicable to the South East Asian region,
prospective study. Ann R Coll Surg Engl 1994; 76:418-9.
which has populations of similar ethnic origins and 6. Baidya N, Rodrigues G, Rao a, Khan SA. Evaluation of Alvarado
diets. The additional parameter of foreign NRIC can score in acute appendicitis: a prospective study. Internet J Surg
[serial online] 2007:9(1). Available at: www.ispub.com/journal/
be included in the score in countries where there is a
the_internet_journal_of_surgery/archive/volume_9_number_
large foreign workforce who has to pay for healthcare 1.html. Accessed May 1, 2008.
treatments. This new appendicitis scoring system is easy 7. Alvarado A. A practical score for the early diagnosis of acute
and simple to apply as the majority of the parameters appendicitis. Ann Emerg Med 1986; 15:557-64.
8. Al-Hashemy AM, Seleem MI. Appraisal of the modified
can be obtained from a routine history and clinical Alvarado Score for acute appendicits in adults. Saudi Med J 2004;
examination. 25:1229-31.
This study was a retrospective analysis of 312 9. Khan I, ur Rehman A. Application of alvarado scoring system
in diagnosis of acute appendicitis. J Ayub Med Coll Abbottabad
patients’ medical records, and hence, the problem of 2005; 17:41-4.
missing data set is a limitation. As shown in Table II, 10. Zweig MH, Campbell G. Receiver-operating characteristic (ROC)
the missing data set ranged from 0.3% to 84%. Despite plots: a fundamental evaluation tool in clinical medicine. Clin
Chem 1993; 39:561-77.
this, the sensitivity and specificity derived for this new
11. Altman DG, Bland JM. Measurement in medicine: the analysis of
appendicitis scoring system, when applied to all 312 method comparison studies. Statistician 1983; 32:307-17.
patients, were comparable to the currently available 12. Kumar V, Cotran RS, Robbins SL. Appendix. In: Robbin’s Basic
Pathology. 5th ed. London:WB Saunders, 1992: 520.
scoring system.
13. Pal KM, Khan A. Appendicitis, a continuing challenge. J Pak Med
The new appendicitis scoring system described in Assoc 1998; 48:189-92.
this study and referred to as the RIPAS Appendicitis 14. Jang SO, Kim BS, Moon DJ. [Application of alvarado score in
patients with suspected appendicitis.] Korean J Gastroenterol
score, or ‘RIPASA’ score in short, is promising and has
2008; 52:27-31. Korean.
good sensitivity, specificity and diagnostic accuracy. 15. Wani MM, Yousaf MN, Khan MA, et al. Usefulness of the
It is simple and easy to use, and has been specifically Alvarado scoring system with respect to age, sex and time of
developed for our local patient group, which is reflective presentation, with regression analysis of individual parameters.
Internet J Surg [serial online] 2007:11(2). Available at: www.
of the South East Asian region in terms of diet and ethnic ispub.com/journal/the_internet_journal_of_surgery/archive/
origin. The prospective evaluation of the RIPASA score volume_11_number_2.html. Accessed July 1, 2008.

You might also like