Cost effectiveness of diagnostic laparoscopy in reproductive aged females suffering from non-specific acute low abdominal pain

Costo efectividad del diagnóstico por laparoscopia del dolor abdominal bajo agudo inespecífico, en mujeres en edad reproductiva

Hernando Gaitán^I; Javier Eslava-Schmalbach^II; Pio Gómez^III

^IMédico. M. Sc. Epidemiología Clínica Clinical Epidemiology Departamento de Ginecología y Obstetricia, Universidad Nacional de Colombia. E-mail: hggaitand@unal.edu.co
^IIMédico. M. Sc. Epidemiología Clínica. Departamento de Cirugía, Instituto de Investigaciones Clínicas, Universidad Nacional de Colombia. E-mail: jheslavas@unal.edu.co ]]> ^IIIMédico. Especialista en Obstetricia y Ginecología Departamento de Ginecología y Obstetricia, Universidad Nacional de Colombia. E-mail: pigomezs@unal.edu.co

ABSTRACT

OBJECTIVE: To compare the costs and effectiveness of early laparoscopy with those of conventional diagnostic methods based on clinical and paraclinical observation and diagnostic images for ascertaining the cause of non-specific acute low abdominal pain (NSALAP) in females of reproductive age from the third-party payers' (TPP) point of view.
METHODS: Population: 110 reproductive aged females suffering from NSALAP. Place: Instituto Materno Infantil, perinatal and maternal attention referral hospital in Bogotá, Colombia. Research design: cost-effectiveness study of a controlled clinical trial carried out in 1998 and 1999. Outcomes to be measured: effectiveness, direct medical costs (in Colombian pesos and their equivalent in US dollars (USD-December 2004) from length of hospital stay, diagnostic procedures carried out, medical visits and managing complications. Analysis: Cost-effectiveness incremental ratio, analysing sensitivity in five different scenarios.
RESULTS: Early diagnostic laparoscopy was more cost-effective in 4 out of the 5 possible scenarios. Savings varying from $21.875 to $69.834 (USD 9.42 and USD 30.1) were made per unit of increased effectiveness.
CONCLUSION: Early diagnostic laparoscopy was cost-effective in 4 out of 5 scenarios dealing with managing NSALAP in reproductive aged females.

Key words: Cost effectiveness, cost analysis, pelvic pain, laparoscopy, women's health (source: DeCS, NLM).

RESUMEN

Palabras-clave: Efectividad, costo y análisis de costo, dolor pélvico, laparoscopia, salud de las mujeres (fuente: DeCS, BIREME).

Acute low abdominal pain is a common reason for reproductive aged females to become hospitalised. 67% of patients suffering from acute abdominal pain present non-specific clinical pictures, this being much more frequent in females (1). A conventional diagnostic approach for determining the cause of pain is based on clinical observation, laboratory tests and diagnostic images (2-6). An alternative approach uses early laparoscopy for making a visual diagnosis of the abdominal-pelvic cavity in an attempt to identify the affected organ (2, 7-9). A situation thus occurs where it becomes necessary to compare two competing diagnostic technologies, understood as such because they are used for the same purpose and are mutually exclusive in the same patient (10), neither one having been shown to be superior over the other in our setting (11)

Each method's effectiveness has to be evaluated when making this comparison; the number of relevant diagnoses, the associated complications and the costs associated with each method must be evaluated. These must be born in mind by society in general, institutions and third-party payers (TPP) since they must become more efficient in handling the scare resources available (12) and provide better quality attention. Laparoscopy represents a case of special attention due to the technique's great spread which has not always been preceded by suitable evaluation.

Our purpose was to carry out an economic analysis from the TPP point of view by comparing both methods' diagnostics costs and effectiveness in terms of dealing with the health of patients suffering from non-specific low acute abdominal (NSALAP), since this investment could lead to greater costs, having equal or less effectiveness, or reduce costs and improve effectiveness.

METHODS

A randomised clinical trial was carried out in the Instituto Materno Infantil between November 1997 and February 2000 for comparing the effectiveness of early laparoscopy (new procedure) and the conventional diagnostic method (standard procedure) in females suffering from NSALAP. This is a referral hospital dealing with maternal, perinatal and gynaecological entities (11).

The costs incurred by both methods in reaching a diagnosis were considered from the TPP point of view. TPP have an important influence on prices paid to health institutions for surgical procedures in the health market in Colombia.

Direct medical costs before diagnosis was made were taken in to account. All costs were related to clarifying the cause of pain. Intra-hospital procedures for making a diagnosis or providing treatment for complications were considered as well as the length of stay associated with them and treatment received which was considered to be unnecessary as it had been based on an erroneous initial diagnosis. The costs of tests which are not routinely used in studying the cause of pain such as endometry biopsy, cervical or endometry cultures were not included, nor were base entity surgical or medical treatment or direct non-medical costs (transport, etc). The number of therapeutic procedures which could be done by laparoscopy avoiding interventions which used to be done by laparotomy was taken into account as being added value. Data were taken from the case report format especially designed for the study and the clinical history. They were estimated from the tariffs for paying the obligatory traffic accident insurance (SOAT). Its tariffs have been homologated for paying for services contracted by the local health entity (Secretaria Distrital de Salud de Bogotá).

The costs at constant prices for 2004 were calculated on tariffs for 1998, adjustments being made for increased consumer price indices for 1998 to 2003 in the following way: a 16,7% increase in 1998, 9,2% for 1999, 8,7% for 2000, 7,6% for 2001, 6,9% for 2002 and 6,4% for 2003. The consumer price index is the basis for readjusting SOAT tariffs in Colombia. A value of 20% per year was taken for calculating depreciation of equipment. No discount was applied since both diagnostic methods were applied for a short pe-riod of time, meaning that the effect of neither could be fully ascertained.

The data for constructing the scenarios when analysing sensitivity were obtained from the available medical literature in relation to length of say, number of diagnostic exams and complications. Expert opinion was consulted when information could not be obtained for the test.

The representative market rate on December 20^th 2004 was taken for estimating the cost in US dollars (i.e. $2 320 Colombian pesos per US dollar).

Defining the terms: Non-specific low abdominal pain: pain about which two observers disagree regarding base diagnosis following six hours' observation or which does not follow a particular disease's classical course; Conventional diagnostic method (standard procedure): diagnosis based on clinical evaluation and diagnostic tests done in parallel or sequentially, the patient being strictly observed. This could include surgical interventions such as precision laparotomy; Laparoscopic method (new procedure): diagnosis based on visualising the abdominal-pelvic cavity and structures contained there, done immediately after the first 6 hours from being admitted and during the first 24 hours of being hospitalised; Diagnostic effectiveness: a method's ability to make an exact diagnosis of a particular disease; Analysis of sensitivity: a strategy used for evaluating the stability of conclusions drawn from analysis when some of the variables (costs, discount rate) change; Scenario: different suppositions constructed for analysing sensitivity (13) using extreme values for a particular variable, obtained from the literature or experts' consensus for evaluating conclusions' stability; Intention to diagnose: analysing the effectiveness of a diagnosis carried out according to patients' initial assignation and taking a particular institution's conditions of daily life into account; Medical direct costs: Hospitalisation costs, Operating room costs for laparoscopy and laparotomy, laboratory costs and those for images and other diagnostic interventions, taking into account the amount paid for lab tests, costs regarding complications.

Indirect or direct non-medical costs such as transport and incapacity were not taken indo account since TPPs do not assume responsibility for paying them in this particular case. Given that only costs for establishing the diagnosis were included, then those produced by treating the base non-complicated entity were not included nor when a complication was present during the first 24 hours of hospitalisation.

The costs above mentioned were taken into account for calculating individual cost (for both the new and standard procedures). 1998 was taken as the base year for estimating costs and effectiveness. Total mean cost per patient and per group was estimated for analysing cost effectiveness.

Mann Whitney U test was used for comparing the costs in each group, given the tendency of data to have an abnormal distribution (14). Intention to diagnose analysis assuming costs of laparotomy in the standard procedure group, even though some patients were subjected to the new procedure for ethical reasons (12).

Cost effectiveness was analysed by cost-effectiveness incremental ratio, obtained by comparing costs/natural outcomes between the groups as follows (15):

Where C_n: new procedure costs, C_c: standard procedure costs.
E_n: new procedure effectiveness, E_c: standard procedure effectiveness.

Incremental cost analysis presented the increase in costs for each increase in effectiveness (15).

Analysis of sensitivity: The two actual scenarios were compared, also the best and worst scenarios for both methods and, the best scenario for the standard procedure compared to the worst scenario for the new procedure and vice versa.

The sensitivity of the effectiveness of both methods was analysed according to the best and worst published scenarios. STATA (version 8.1) software was used for statistical analysis.

RESULTS

110 females suffering from non-specific acute low abdominal pain were admitted between November 1997 and February 2000. Table 1 shows patients' base characteristics and final diagnosis. Laparoscopy provided 82% correct diagnoses; standard procedure was 78% effective (11).

Table 1. Base characteristics and final diagnosis in 110 females suffering
from non-specific pain in the Instituto Materno Infantil - Bogotá, Colombia

Table 2 shows the distribution of resources used for achieving a final diagnosis. Length of stay needed the greatest amount of resources, the greatest number of medical visits and the greatest number of haemograms and ecographies were requested for the standard procedure group.

Table 2. Length of stay, medical visits and laboratory exams according
to diagnostic method in patients suffering non-chronic pelvic pain in the
Instituto Materno Infantil - Bogota, Colombia

Mean ± standard deviation; Number (%); Median (range);
* T-Student test was used for normal distribution variables;
** Mann Whitney U test was used for abnormal distribution variables.
Significance level: 0.01; Number of patients related to outcome (percentage)

Sensitivity Analysis: The best scenario when analysing cost sensitivity for the standard procedure was considered to be a one-day stay, just base ex-ams being done and that there was no complication requiring laparotomy or uterine curettage. The series having the least number of entities presenting NSALAP revealed a series of 41 patients, having unexplained abdominal pain, 5% of them requiring surgery (16).

The worst scenario for the standard procedure considered a 4,1 day-stay for patients having non-specific abdominal pain described in a series of 100 patients (1) Regarding laboratory tests or diagnostic images, it was assumed that three haemograms, two uroanalyses, two qualitative pregnancy tests and two pelvic and transvaginal ecographies would be requested for the longest stay.

The greatest numbers of patients possibly requiring surgery were found ranging from 41%, in a series of 100 cases of patients having suspected appendicitis or gynaecological pathology, surgery was performed on 41 patients (17), to 62% in a series of 119 patients suffering from acute abdomi-nal pain studied (18). The frequency given by the former publication was chosen as its population was similar that being studied here.

Regarding complications, a 10% frequency in delay was determined for establishing secondary complications as being the cause of pain. This figure was obtained by considering a 6% frequency of perforated appendicitis. The foregoing must be added to 1,5% of patients having ectopic pregnancy could present complications and 2,5% of patients having slight EPI could present secondary complications when diagnosis is delayed. Such a figure is similar to that referred to by Nevez in patients suffering from acute abdominal pain (19). An average of two days' prolonged hospital stay was estimated and that 4 out of each 5 patients having complications would require open surgery for treating their complications.

The best scenario for the laparoscopic method considered that there would be no hospitalisation costs because a patient is treated as an outpatient, bearing in mind that a patient presenting a gynaecological pathology can be handled as an outpatient. Only initial exams would be done, no complications would arise and that up to 85% of major surgery could be avoided if operation laparoscopy were done immediately.

For constructing the new procedure's worst scenario.was obtained by considering a two-day stay for post-operation care taken from a study published on a series of 19 patients (20). A 10% complication frequency reported (19), with 20% laparotomy frequency was considered, since laparoscopy did not provide sufficient information (21), without which major surgery could be avoided due to technical limitations, particular norms applying to a particular service and the low frequency of a pathology requiring therapeutic surgical procedures. Regarding complications, it was considered that all patients presenting serious complications would require laparotomy for completing the procedure and examining the cavity for discarding vascular or visceral complications. It was also considered that hospitalisation would be prolonged by two days for managing such complication.

Tables 3a and 3b shows hospital costs in Colombian pesos and US dollars (US$) for each procedure related to average stay, lab test and pelvic ultrasounds used in each method, the number of surgical interventions carried out in each group and procedures avoided, as well as the number of procedures in the best and worst scenarios. Total average costs per patient in each group were obtained from this data.

Table 3b. Real estimation of costs for different procedures according for
Laparoscopic diagnostic method, used in a series of 110 patients suffering from
non-specific low abdominal pain in the Instituto Materno Infantil - Bogotá, Colombia

Overall comparison of costs between both methods revealed no statistically significant differences in the real scenario (p = 0,74 Mann Whitney U test); however, lower hospital costs were observed in the group of patients submitted to laparoscopy. The cost-effectiveness incremental ratio, of the two methods being studied was analysed in the different scenarios, taking the foregoing information into account.

Scenario 1: actual scenario, data obtained from the study.

Laparoscopy would be less expensive and more effective. The TPP would save $25 792 per patient (USD 11,1) for each unit of diagnostic effectiveness increasing with laparoscopy.

Scenario 2: the best scenario for both methods:

Laparoscopy would be less expensive and more effective. The TPP would save $ 69 834 ($USD 30,1) for each unit of diagnostic effectiveness increasing with laparoscopy.

Scenario 3: the worst scenario for both methods:

Laparoscopy would be less expensive and more effective. The TPP would save $ 21 875 (US$ 9,4) by using laparoscopy for checking a diagnosis, per increase in each unit of effectiveness.

Scenario 4: the best scenario for new procedure cf worst standard procedure scenario:

Laparoscopy would be less expensive and more effective. The TPP would save $30 006 (US$ 12,9) by using laparoscopy for checking a diagnosis, for each increased unit of effectiveness.

Scenario 5: worst new procedure scenario cf best standard procedure scenario:

Laparoscopy would be more costly and less effective.. The hospital would have been paid $35 696 (USD 15,3) more for each unit of reduced effectiveness in this situation.

The new procedure would be more cost-effective than the standard procedure for patients suffering from NSALAP in two scenarios. This was equally effective in two scenarios but led to minimising costs in the study of patients suffering NSALAP. The new procedure was not cost-effective in one scenario since it reduced effectiveness at greater cost.

The economic study of implementing programmes for attention, diagnosis or therapeutic intervention has become a current need given health systems' limited resources whether provided by government or private agencies.

This type of approach provides another tool for evaluating the performance of strategic alternatives, especially when they are mutually exclusive since applying one method doe not allow the other to be applied, as is the case when applying early endoscopic diagnostic technology.

Diagnostic technologies could be evaluated from the point of view of their diagnostic validity (sensitivity and specificity), reliability (intra- or inter-observer agreement) and effectiveness, understood as being the suitable use of an intervention according to a specific situation (22). In this case, it would be a technology's performance for diagnostic use and the cost effectiveness or analysis of the cost of implementing this new technology related to the increase in expected or acceptable minimum exactitude given this new application of technology.

Cost-effectiveness studies can be based on information obtained from revising the literature for constructing some theoretical scenarios or based on controlled clinical trials (CCT) evaluating the effectiveness provided by data emerging from the real scenario of daily life in the area of work and resort to secondary sources of information in the literature for assembling the best and worst scenarios in which the technology to be evaluated is to be analysed (13). The methodology based on CCT is used in this study, providing greater validity for the estimates made.

Their application is limited since diagnostic evaluation methods depend on institutional protocols and the tacit knowledge at hospitals' disposal, known as "the medical school", meaning that this could vary the number of exams requested. Other factors limiting results being generalised are the curve of experience, the degree of available technological development and the value of services which are contracted by TPPs.

However, contracted services can be calculated from tariffs recognised by each TPP in each institution, based on information provided regarding aspects such as the number of procedures carried out, days being hospitalised and medical visits required. They do allow modifications to be introduced to best and worst performance scenarios, given the characteristics common to an institution with which it is wished to contract.

As this evaluation was made from a TPP point of view, it could have been desirable to estimate indirect costs such as the cost of incapacity and reduced productivity; however, this was beyond the scope of the present study.

This study showed that the new procedure was more cost-effective (at SOAT tariffs) than the standard procedure for studying reproductive aged females suffering from non-specific abdominal pain from the TPP point of view in 4 out of 5 scenarios. This study thus provides useful information for decision-makers for helping to implement new programmes

1. Sheridan WG, White AT, Havard T, Crosby DL. Non-specific abdominal pain: the resource implications. Ann R Coll Surg Engl 1992;74:181-185.        [ Links ]

2. Sellors J, Mahony J, Goldsmith C, Rath D, Mander R, Hunter B, et al. The accuracy of clinical findings and laparoscopy in pelvic inflammatory disease. Am J Obstet Gynecol 1991;164:113-20        [ Links ]

3. Brown DL, Doulbilet PM. Transvaginal sonography for diagnosing ectopic pregnancy. Positive criteria and performance characteristics. L Ultrasound Med 1994;13:259-266        [ Links ]

4. Hoffman J, Rasmussen O. Aids in the diagnosis of acute appendicitis. Br. J. Surgery 1989;76:774-779.        [ Links ]

5. Olsen JB, Myrén CJ, Haahr PE. Randomized study of the value of laparoscopy before appendicectomy. Br. J. Surgery 1993;80:922-923.        [ Links ]

6. Sarfati MR, Hunter GC, Witzke DB, Bebb GG, Smythe SH, Boyan et al. Impact of adjunctive testing on the diagnosis and clinical course of patients with acute appendicitis. Am J of Surgery 166:660-665.        [ Links ]

7. Spirtos NM, Eisenkop SM, Spirtos TW, Poliakin RI, Hibbard LT.. Laparoscopy. A diagnosis aid in cases of suspected appendicitis. Its use in women of reproductive age. Am. J. Obstet Gynecol 1987;156:90-94.        [ Links ]

8. Decadt B, Sussman L, Lewis MP, Secker A, Cohen L, Rogers C, et al. Randomized clinical trial of early laparoscopy in the management of acute non-specific abdominal pain. Br. J. Surg 1999;86:1383-1386.        [ Links ]

9. Llanio R. Laparoscopia de urgencias. 1^st edition, La Habana, Editorial Científico Técnico, 1977.        [ Links ]

10. Petitti DB. Competing technologies. Implications for costs and complexity of medical care. N. Eng. J. Med 1986;315:1480-1483.        [ Links ]

11. Gaitán H, Angel E, Sanchez J, Gomez I, Sanchez L, Agudelo C. Laparoscopic diagnosis of acute lower abdominal pain in women of reproductive age. Int J Gynaecol Obstet 2002 Feb;76(2):149-58.        [ Links ]

12. Law 100 /1993. Concerning the Colombian social security system, República de Colombia.        [ Links ]

13. Petitti, D. Overview of the methods, Chapter 2, in Petitti D. Meta-analysis, decision analysis and cost effectiveness analysis. Methods for quantitative synthesis in medicine. 2^nd edition, Oxford University Press, New York, 2000, p.30-31).        [ Links ]

14. Diehr P, Yanez D, Ash A, Hornbrook M, Lin DY.. Methods for analyzing health care utilization and costs. Annu Rev Public Health 1999;20:125-144.        [ Links ]

15. Petitti DB. Advanced cost-effectiveness analysis Chapter 12. In Petitti DB. Meta-analysis, decision analysis and cost-effectiveness analysis. 1^st edition, Oxford University Press, New York, 1994.        [ Links ]

16. Champault G, Lauroy J, Guillon P, Benoit J, Rizk N, Boutelier J. What are abdominal painful syndromes of unexplained origin? Prospective study: 99 patients followed for three years]. Ann Chir. 1996;50(3):258-62.        [ Links ]

17. Rao PM, Feltmate CM, Rhea JT, Schulick AH, Novelline RA.. Helical computed tomography in differentiating appendicitis and acute gynecological conditions. Obstet Gynecol 1999;93:417-421.        [ Links ]

18. Salky BA, Edye MB. The role of laparoscopy in the diagnosis and treatment of abdominal pain syndromes. Surgical Endoscopy 1998;12:911-914.        [ Links ]

19. Navez B, d'Udekem Y, Cambier E, Richir C, de Pierpont B, Guiot P. Laparoscopy for management of non traumatic acute abdomen. World J Surg 1995;19:382-386.        [ Links ]

20. Taylor EW, Kennedy CA, Dunham RH, Bloch JH. Diagnosis laparoscopy in women with acute abdominal pain. Surg Laparosc Endosc 1995;5:125-128.        [ Links ]

21. Cuesta MA, Eijsbouts QA, Gordijn RV, et al. Diagnosis laparoscopy in patients with acute abdomen of uncertain etiology. Surg Endosc 1998;12:915-917.        [ Links ]

22. Roper WL, Winkenwerder W, Hackbarth GM, Krakauer H. Effectiveness in health care. An initiative to evaluate and improve medical practice. N Engl J Med 1988;319:1197-202.        [ Links ]

Recibido 31 Enero 2005 / Enviado para Modificación 25 Febrero 2005 / Aceptado 8 Junio 2005