Bruno Dujardin

Likelihood ratios: A real improvement for clinical decision making?

The concept of likelihood ratio has been advocated for several years as one of the better means to evaluate diagnostic tests and as a practical and valuable tool in clinical decision making. In this paper we review the basic concepts underlying the evaluation of diagnostic tests and we explore the properties and usefulness of both positive and negative likelihood ratios compared with sensitivity and specificity. Particular attention is given to the use of likelihood ratios in the clinical setting. Likelihood ratios have three main advantages: they are intuitive, they simplify the predictive value calculation and the overall evaluation of sequential testing. Disadvantages are the non-linearity and the necessity to recalculate probabilities in odds. Although they summarize the information contained in sensitivity and specificity, these characteristics are still necessary for certain clinical decisions. Since likelihood ratios have been promoted among physicians and medical students, we discuss examples of inappropriate use and misunderstandings in the medical literature: the frequent omission of confidence intervals, the choice of cut-off points based on likelihood ratios for positive test results only and the confusion between likelihood ratios for ranges and those for cut-off points.

The strategy of risk approach in antenatal care: Evaluation of the referral compliance

The main goal of antenatal care in developing countries is to identify women whose pregnancy or delivery is likely to raise problems and to refer them at the appropriate time to a hospital facility where the necessary medical equipment and expertise (vacuum extractors, cesarian sections, human skill, etc.) is available. This approach, which is known as the Risk Approach (RA) strategy, is expected to significantly reduce maternal morbidity and mortality. However, the RA will function properly only if the women identified at risk agree to give birth in a hospital on the one hand, and if they can indeed reach this hospital on the other hand. In this article the authors assess to what extent women with a risk of difficult labor (nulliparous or primiparous women under 150 cm, history of previous difficult delivery or stillbirth, women with transverse lie) agreed to give birth in a hospital. This descriptive survey, which covered 5060 pregnancies monitored in the Kasongo District, Maniema, in eastern Zaire, showed that the referral success rate in this socioeconomically very disadvantaged region was only 33%, despite some favorable conditions, such as a strong emphasis on community participation, a complementarity of health centers and hospital, and the absence of financial barriers within the health services system. Of the various hypotheses tested, the geographic accessibility of the hospital and the parturients perception of the risk status were the two most important factors determining the compliance rate. A stratified analysis shows that the intensity of the parturients perception has a different impact on compliance whether rural or urban situations are considered.(ABSTRACT TRUNCATED AT 250 WORDS)

The value of maternal height as a risk factor of dystocia: a meta‐analysis

Ten publications and studies on the relation between maternal height and the risk of dystocia due to cephalopelvic disproportion (CPD) are analysed. The rate of Caesarean sections was chosen as the CPD indicator. When maternal height is presented in percentiles, curves can be superimposed, and sensitivities and specificities of the various studies may be analysed together. One biased study was excluded; the remaining 9 were pooled and regression lines calculated for sensitivity (Se) and specificity (Sp) of the entire set of points. The resulting model, i.e. Se = 10.9+1.99 Y and Sp = 99.9 − 0.99 Y, permits easy calculation of the expected sensitivity and specificity for each percentile Y. When the frequency of Caesarean section due to CPD is known, positive and negative predictive values can also be calculated. The proposed formulas can also be used to determine confidence intervals.

Oxytocics in developing countries

Objectives: A prospective multicenter study of obstetric practices was conducted in three developing countries (Benin, Congo and Senegal) to analyze oxytocic use during labor. One of the objectives was to assess the possible negative effects of the treatment regimens instituted during the labor monitoring phase. Methods: Four health districts participated in the study. All women who gave birth in one of the participating health facilities over a 6‐month period in Benin and Congo, and over a 3‐month period in Senegal, were recruited. The number of deliveries studied in each district varied from 457 to 1048. For each case a partogram was used to assess the progress of labor and the onset of dysfunctional labor. Information was collected on the risk factors for dysfunctional labor, stillbirths and resuscitation of the neonate. Results: Each of the four collaborating centers used oxytocics preferentially to treat dysfunctional labor, but even in normal labor (i.e. with a normal partogram) oxytocics were used in 4.4–21.5% of cases. In normal labor the incidence of neonatal resuscitation was higher in cases with than in those without oxytocic use: the relative risks (R.R.) varied from 1.9 to 5.6; the odds ratios varied from 2.4 to 7.0, and both were statistically significant in the four settings. In addition the stillbirth rate was always higher, though not significantly, when oxytocics were used in normal labor (R.R. 1.2–2.2). When the data of the four centers were pooled, the global relative risk for stillbirths was 1.9, and the 95% confidence interval was 1.1–3.4. Logistic regression analysis was carried out for five confounding factors (primiparity, a previous complicated delivery, presence of meconium, ruptured membranes and educational level) to adjust the odds ratio for the risk of neonatal resuscitation when oxytocics were used in normal labor. Except in the case of Abomey in Benin, where the variable ‘presence of meconium’ decreased the odds ratio from 6.4 to 3.4, the adjusted odds ratios remained similar to their non‐adjusted values. In cases of non‐dysfunctional labor, nurses and midwives used oxytocics more often than lesser trained health personnel (R.R. 4.0 [3.2–5.1]). Conclusion: Our results show that an obstetric treatment which is safe when used in certain well‐defined indications, may have significant negative effects when used in situations where the same technical quality of care cannot be guaranteed.

Editorial: Tuberculosis control: did the programme fail or did we fail the programme?

Under pressure of the increasing numbers of tuberculosis (TB) cases in the world, TB control has once again become a major challenge. As such it is the subject of intensive scientific activity, as evidenced by the numerous studies and publications that have been devoted to it over the last few years. The Lancet recently published two documents which summarize present concerns: the minutes of the conference organized in Washington DC by The Lancet, in September 1995: ‘The challenge of tuberculosis: statements on global control and prevention’ (Enarson et al. 1995) and an article which proposes substantial modifications of activities in the standard tuberculosis control programme (De Cock & Wilkinson 1995). From the most recent literature, two major challenges may be identified (Reichman & Hershfield 1993; Porter & McAdam 1994; De Cock & Wilkinson 1995; Enarson et al. 1995): On the one hand there is a call for the development of new diagnostic techniques, especially procedures that are faster and more sensitive than smears or cultures and techniques that would improve or facilitate the diagnosis of smearnegative TB; and a call for new treatments that are effective against multidrug-resistant TB and/or that would shorten length of treatment. On the other hand, the scientific community also acknowledges the importance of some operational aspects of TB, such as problems of drugs delivery and financing, and patient compliance to treatment (Reichman & Hershfield 1993; Porter & McAdam 1994; De Cock & Wilkinson 1995; Enarson et al. 1995). This last point is considered a top priority, and WHO is currently promoting DOT (Daily Observed Therapy) as a new strategy to be implemented by each TB control programme (Enarson et al. 1995). However, other aspects linked to the organization and the functioning of health services, or linked to the perception of the illness by both health personnel and patients, are underestimated. In his presidential address, given at the 21st Andhra Pradesh TB and Chest Diseases Conference held in July 1994 in India, Dr Ranga Rao proposed a critical self-evaluation of the state TB control programme which started more than three decades ago (Rao 1994). This physician, who has been working as a TB officer for more than 25 years, identified 17 major weaknesses of the TB control programme. His very impressive list begins with:

Validité, coût et faisabilité de la mAECT et CTC comme tests de confirmation dans la détection de la Trypanosomiase Humaine Africaine

Objectifs  Evaluer la validité, le coût et la faisabilité de deux tests parasitologiques pour la confirmation de la maladie du sommeil; la mini Anion Exchange Centrifugation Technique (mAECT) et la Capillary Tube Centrifugation (CTC).

Cost-effectiveness of algorithms for confirmation test of human African trypanosomiasis

Algorithms that incorporate concentration techniques are more effective and efficient than the currently used algorithms.

[Validity, cost and feasibility of the mAECT and CTC confirmation tests after diagnosis of African of sleeping sickness].

Objectifs  Evaluer la validité, le coût et la faisabilité de deux tests parasitologiques pour la confirmation de la maladie du sommeil; la mini Anion Exchange Centrifugation Technique (mAECT) et la Capillary Tube Centrifugation (CTC).

Adherence to miltefosine treatment for visceral leishmaniasis under routine conditions in Nepal.

To assess patient adherence to unsupervised single‐drug miltefosine treatment for visceral leishmaniasis and to identify the factors influencing adherence.

Determinants of Maternal Near-Miss in Morocco: Too Late, Too Far, Too Sloppy?

Background In Morocco, there is little information on the circumstances surrounding maternal near misses. This study aimed to determine the incidence, characteristics, and determinants of maternal near misses in Morocco. Method A prospective case-control study was conducted at 3 referral maternity hospitals in the Marrakech region of Morocco between February and July 2012. Near-miss cases included severe hemorrhage, hypertensive disorders, and prolonged obstructed labor. Three unmatched controls were selected for each near-miss case. Three categories of risk factors (sociodemographics, reproductive history, and delays), as well as perinatal outcomes, were assessed, and bivariate and multivariate analyses of the determinants were performed. A sample of 30 near misses and 30 non-near misses was interviewed. Results The incidence of near misses was 12‰ of births. Hypertensive disorders during pregnancy (45%) and severe hemorrhage (39%) were the most frequent direct causes of near miss. The main risk factors were illiteracy [OR = 2.35; 95% CI: (1.07–5.15)], lack of antenatal care [OR = 3.97; 95% CI: (1.42–11.09)], complications during pregnancy [OR = 2.81; 95% CI:(1.26–6.29)], and having experienced a first phase delay [OR = 8.71; 95% CI: (3.97–19.12)] and a first phase of third delay [OR = 4.03; 95% CI: (1.75–9.25)]. The main reasons for the first delay were lack of a family authority figure who could make a decision, lack of sufficient financial resources, lack of a vehicle, and fear of health facilities. The majority of near misses demonstrated a third delay with many referrals. The women’s perceptions of the quality of their care highlighted the importance of information, good communication, and attitude. Conclusion Women and newborns with serious obstetric complications have a greater chance of successful outcomes if they are immediately directed to a functioning referral hospital and if the providers are responsive.