The effect of new Mycobacterium tuberculosis infection on the sensitivity of prognostic TB signatures

BACKGROUND: Tests that identify individuals at greatest risk of TB will allow more efficient targeting of preventive therapy. The WHO target product profile for such tests defines optimal sensitivity of 90% and minimum sensitivity of 75% for predicting incident TB. The CORTIS (Correlate of Risk Targeted Intervention Study) evaluated a blood transcriptomic signature (RISK11) for predicting incident TB in a high transmission setting. RISK11 is able to predict TB disease progression but optimal prognostic performance was limited to a 6-month horizon.METHODS: Using a mathematical model, we estimated how subsequent Mycobacterium tuberculosis (MTB) infection may have contributed to the decline in sensitivity of RISK11. We calculated the effect at different RISK11 thresholds (60% and 26%) and for different assumptions about the risk of MTB infection.RESULTS: Modelled sensitivity over 15 months, excluding new infection, was 28.7% (95% CI 12.3-74.1) compared to 25.0% (95% CI 12.7-45.9) observed in the trial. Modelled sensitivity exceeded the minimum criteria (>75%) over a 9-month horizon at the 60% threshold and over 12 months at the 26% threshold.CONCLUSIONS: The effect of new infection on prognostic signature performance is likely to be small. Signatures such as RISK11 may be most useful in individuals, such as household contacts, where probable time of infection is known.

Seasonal drivers of tuberculosis: evidence from over 100 years of notifications in Cape Town

BACKGROUND: Tuberculosis incidence varies seasonally in many settings. However, the role of seasonal variation in reactivation vs. transmission is unclear.METHODS: We reviewed data on TB notifications in Cape Town, South Africa, from 1903 to 2017 (exclusive of 1995-2002, which were unavailable). Data from 2003 onward were stratified by HIV status, age and notification status (new vs. retreatment). We performed seasonal decomposition and time-dependent spectral analysis using wavelets to assess periodicity over time. We estimated monthly peak-to-peak seasonal amplitude of notifications as a percentage of the annual notification rate.RESULTS: A seasonal trend was intermittently detected between 1904 and 1994, particularly during periods of high notification rates, but was consistently and strongly evident between 2003 and 2017, with peaks in September through November, following winter. Among young children, a second, higher seasonal peak was observed in March. Seasonal variation was greater in children (<5 years, 54%, 95% CI 47-61; 5-14 years, 63%, 95% CI 58-69) than in adults (36%, 95% CI 33-39).CONCLUSIONS: Stronger seasonal effects were seen in children, in whom progression following recent infection is known to be the predominant driver of disease. These findings may support increased transmission in the winter as an important driver of TB in Cape Town.

Temporal trends in the prevalence of Mycobacterium tuberculosis infection in South African adolescents

<sec id="st1"> <title>SETTING</title> South Africa. </sec> <sec id="st2"> <title>OBJECTIVE</title> 1) To measure changes in the adolescent prevalence of latent tuberculous infection (LTBI) between 2005 and 2015, and 2) to evaluate medium-term impact of TB control measures on LTBI prevalence. </sec> <sec id="st3"> <title>DESIGN</title> We compared baseline data from a cohort study (2005-2007) and a vaccine trial (2014-2015) which enrolled adolescents from the same eight South African high schools. LTBI was defined based on QuantiFERON^®-TB Gold In-Tube test positivity. </sec> <sec id="st4"> <title>RESULTS</title> We analysed data from 4880 adolescents between 2005 and 2007, and 1968 adolescents between 2014 and 2015, when the average LTBI prevalence was respectively 43.8% (95%CI 28.4-59.1) vs. 48.5% (95%CI 41.1-55.8). Age-specific LTBI prevalence increased between the ages 12 and 18 years by 13% only in lower socio-economic quintile schools, where the average LTBI prevalence was unchanged between the two periods (54% vs. 53%). In the highest socio-economic quintile schools, LTBI prevalence did not increase with age; however, the average LTBI prevalence increased from 20% to 38% between the two periods. </sec> <sec id="st5"> <title>CONCLUSION</title> Adolescent LTBI prevalence remained high and constant over a decade, suggesting that Mycobacterium tuberculosis transmission to children was not impacted in the medium term by effective TB control efforts. Trends in adolescent LTBI prevalence should be interpreted in the context of the sociodemographic factors that affect the risk of transmission before and during adolescence. </sec>.

Impact of isoniazid preventive therapy on the evaluation of long-term effectiveness of infant MVA85A vaccination

SETTING:

South Africa.

OBJECTIVE:

To evaluate the long-term effectiveness of infant modified vaccinia Ankara virus-expressing antigen 85A (MVA85A) vaccination against tuberculosis (TB).

DESIGN:

We analysed data from a double-blind randomised placebo-controlled Phase 2b MVA85A infant TB vaccine trial (2009–2012), with extended post-trial follow-up (2012–2014). Isoniazid preventive therapy (IPT) was provided by public health services according to national guidelines. The primary outcome was curative treatment for TB disease. Survival analysis and Poisson regression were used for study analysis.

RESULTS:

Total follow-up was 10 351 person-years of observation (pyo). Median follow-up age was 4.8 years (interquartile range 4.4–5.2). There were 328 (12%) TB cases. TB disease incidence was 3.2/100 pyo (95%CI 2.8–3.5) overall, and respectively 3.3 (95%CI 2.9–3.9) and 3.0 (95%CI 2.6–3.5)/100 pyo in the MVA85A vaccine and placebo arms. A total of 304 children (11%) received IPT, with respectively 880 and 9471 pyo among IPT and non-IPT recipients. There were 23 (7.6%) TB cases among 304 IPT recipients vs. 305 (12.9%) among 2374 non-IPT recipients (P = 0.008). IPT effectiveness was 85% (95%CI 76–91).

CONCLUSION:

Extended follow-up confirms no long-term effectiveness of infant MVA85A vaccination, but a six-fold reduction in TB risk can be attributed to IPT. National TB programmes in high TB burden countries should ensure optimal implementation of IPT for eligible children.

Comparison of haematology and biochemistry parameters in healthy South African infants with laboratory reference intervals

Objective

Paediatric laboratory reference intervals used in Africa and Asia may be derived from historical intervals of predominantly Caucasian infants in Europe or North America. These intervals may therefore not be compatible with the range of normality for developing country populations. We aimed to compare haematology and biochemistry parameters in healthy South African infants with local laboratory reference intervals.

Methods

We compared the baseline haematology and biochemistry results of 634 (316 male and 318 female) HIV‐unexposed infants, aged 3–6 months, living in a rural area of the Western Cape Province, South Africa, against laboratory reference intervals supplied by the South African National Health Laboratory Services (NHLS). We calculated the percentage of observed values out of bound (in terms of lower and upper limits) compared to laboratory reference intervals.

Results

Of the 634 healthy infants screened, 316 (49.84%) were male and 318 (50.16%) female. A majority (91.05%) had platelet counts above the laboratory reference interval upper limit (350 × 10⁹cells/l), while over half, 54.85% and 56.98% had mean corpuscular volume (MCV) and mean corpuscular haemoglobin (MCH) values below the lower limits of 77.0–105.0 fl and 26.0–34.0 pg, respectively. A small proportion were outside the reference limits for haematocrit, namely 15.71% below and 7.14% above the normal limits of 0.31–0.38 l/l. For male and female infants, 33.65% and 18.04% of alkaline phosphatase (ALP) values and 7.01% and 14.56% of alanine transaminase (ALT) values were above the upper limits, respectively. For male infants, 10.83% of gamma‐glutamyl transferase (GGT) values, and for female infants, 31.11% of GGT values were below the lower limits of 12 U/l for males and 15 U/l for females. We observed no significant deviations (>10% out of bound) from NHLS reference intervals in the remaining haematology and biochemistry parameters measured.

Conclusions

Haematology and biochemistry parameters in apparently healthy South African infants deviate frequently from national laboratory reference intervals, including abnormalities consistent with subclinical hypochromic microcytic anaemia. It is important that clinical laboratory reference intervals for children are derived locally, rather than being adopted from Caucasian norms in developed countries, because clinical trials of vaccines, drugs and diagnostics are increasingly conducted in sub‐Saharan Africa.

A controlled trial of sputum induction and routine collection methods for TB diagnosis in a South African community

The diagnostic yield of pulmonary tuberculosis (TB) by sputum induction (SI) at the first point of contact with health services, conducted in all patients with suspected TB regardless of the ability to expectorate spontaneously, has not been evaluated. We compared the diagnostic yield of SI to routine sputum collection in a South African community setting. Ambulatory patients with suspected TB provided a 'spot' expectorated sputum sample, an SI sample by hypertonic (5 %) saline nebulization, and early morning expectorated sputum sample. The diagnostic yield of sputum smear microscopy and liquid culture (denominator all subjects with any positive Mycobacterium tuberculosis culture), and time-to-positivity of culture were compared between SI and expectorated samples. A total of 555 subjects completed the SI procedure, of whom 132 (24 %) were human immunodeficiency virus (HIV)-infected. One hundred and twenty-nine samples (129, 23 %) were M. tuberculosis culture-positive. The time-to-positivity of Mycobacteria Growth Indicator Tube (MGIT) culture was shorter for SI (median difference 2 days, p = 0.63) and for early morning expectorated sputum (median difference 2 days, p = 0.02) compared to spot expectorated sputum. However, there was no difference in the culture-positive diagnostic yield between SI and spot expectorated sputum [difference +0.7 %; confidence interval (CI) -7.0 to +8.5 %, p = 0.82] or SI and early morning expectorated sputum (difference +4.7 %; CI -3.2 to +12.5 %, p = 0.20) for all subjects or for HIV-infected subjects. SI reduces the time to positive M. tuberculosis culture, but does not increase the rate of positive culture compared to routine expectorated sputum collection. SI cannot be recommended as the routine collection method at first contact among ambulatory patients with suspected TB in high-burden communities.

Risky behaviour and psychosocial correlates in adolescents - is there a link with tuberculosis?

OBJECTIVE

Reasons for the increase in incidence of tuberculosis (TB) in late adolescence are poorly understood. One hypothesis is that psychological and behavioural variables associated with adolescence may increase risk of developing TB. The study aimed to determine whether psychosocial and behavioural variables affect incidence of TB disease in adolescents.

METHODS

A case-control study design was used in adolescents who were participants in a TB epidemiological study. Cases were adolescents diagnosed with TB disease. Approximately half of the controls had no TB disease but a positive TST indicative of latent TB. Half had neither TB disease nor latent TB. A self-administered questionnaire was completed by participants. The questionnaire consisted of a combination of standardised psychosocial instruments.

RESULTS

Of 292 participants, 62 were cases, 112 had latent TB and 118 neither TB disease nor latent TB. There were no significant differences in instrument scores between cases and controls. There was a trend for certain adverse life events to be more common in the TB-disease group.

CONCLUSION

In adolescents, a trend for association between TB incidence and psychosocial and behavioural variables was not statistically significant. Given the trend, research with larger samples, and more comprehensive assessment of the relationship between stressors and TB, is warranted.

Vaccine trials in the developing world: operational lessons learnt from a phase IV poliomyelitis vaccine trial in South Africa

BACKGROUND

Conducting vaccine trials in developing nations is necessary but operationally complex. We describe operational lessons learnt from a phase IV poliomyelitis vaccine trial in a semi-rural region of South Africa.

METHODS

We reviewed operational data collected over the duration of the trial with respect to staff recruitment and training, participant recruitment and retention, and cold chain maintenance. RESULTS-LESSONS LEARNT: The recruitment model we used that relied on the 24h physical presence of a team member in the birthing unit was expensive and challenging to manage. Forecasting of enrolment rates was complicated by incomplete baseline data and by the linear nature of forecasts that do not take into account changing variables. We found that analyzing key operational data to monitor progress of the trial enabled us to identify problem areas timeously, and to facilitate a collegial problem-solving process by the extended trial team. Pro-actively nurturing a working relationship with the public sector health care system and the community was critical to our success. Despite the wide geographical area and lack of fixed addresses, we maintained an excellent retention rate through community assistance and the use of descriptive residential information. Training needs of team members were ongoing and dynamic and we discovered that these needs that were best met by an in-house, targeted and systemized training programme. The use of vaccine refrigerators instead of standard frost-free refrigerators is cost-effective and necessary to maintain the cold-chain.

CONCLUSION

Operational challenges of a vaccine trial in developing world populations include inexperienced staff, the close liaison required between researchers and public health care services, impoverished participants that require complex recruitment and retention strategies, and challenges of distance and access. These challenges can be overcome by innovative strategies that allow for the unique characteristics of the setting, trial population, and trial team.

Tuberculosis case finding for vaccine trials in young children in high-incidence settings: a randomised trial

SETTING

A high tuberculosis (TB) burden rural area in South Africa.

OBJECTIVE

To compare TB case yield and disease profile among bacille Calmette-Guérin (BCG) vaccinated children using two case-finding strategies from birth until 2 years of age.

DESIGN

BCG-vaccinated infants were enrolled within 2 weeks of birth and randomised to 3-monthly home visits for questionnaire-based TB screening plus record surveillance of TB registers, hospital admission and X-ray lists at health facilities for TB suspects and cases (Group 1), or record surveillance (as above) only (Group 2). Both groups received a close-out visit after 2 years. Participants were evaluated for suspected TB disease using standardised investigations.

RESULTS

A total of 4786 infants were enrolled: 2392 were randomised to Group 1 and 2394 to Group 2. The case-finding rate was significantly greater in Group 1 (2.2/100 py) than in Group 2 (0.8/100 py), with a case-finding rate ratio of 2.6 (95%CI 1.8-4.0, P < 0.001). Although the proportion of cases with bacteriological confirmation was lower in Group 1, this difference did not reach statistical significance. There was also no significant difference in the proportions with TB symptoms and signs.

CONCLUSION

Home visits combined with record surveillance detected significantly more cases than record surveillance with a single study-end visit. The TB case profile did not differ significantly between the two groups.

Tuberculin skin test and QuantiFERON® assay in young children investigated for tuberculosis in South Africa

SETTING

Although the literature on interferon-gamma release assays on tuberculosis (TB) in children has increased, data pertaining to young children remain relatively limited.

OBJECTIVE

To compare results from the tuberculin skin test (TST) and the QuantiFERON®-TB Gold In-Tube assay (QFT) in children aged <3 years investigated for TB disease.

DESIGN

TB suspects were evaluated by medical history and examination, TST, QFT, chest radiography, induced sputum and gastric washings for smear and culture for Mycobacterium tuberculosis.

RESULTS

A total of 400 children were enrolled. Among 397 children with both test results, 68 (17%) were QFT-positive and 72 (18%) were TST-positive (≥10 mm). Agreement between the tests was excellent (94%, κ = 0.79, 95%CI 0.69-0.89). TB disease was diagnosed in 52/397 (13%) participants: 3 definite, 35 probable and 14 possible TB. QFT sensitivity and specificity for TB disease were respectively 38% and 81%. TST sensitivity and specificity were respectively 35% and 84%.

CONCLUSION

While TST and QFT had excellent concordance in this population, both tests had much lower sensitivity for TB disease than has been reported for other age groups. Our results suggested equivalent performance of QFT and TST in the diagnosis of TB disease in young children in a high-burden setting.

Sputum induction for the diagnosis of pulmonary tuberculosis: a systematic review and meta-analysis

Sputum induction (SI) has been proposed as the optimal sample collection method for patients with paucibacillary tuberculosis (TB). Studies reporting the culture of Mycobacterium tuberculosis from SI were reviewed. A random-effects meta-analysis of diagnostic yield (numerator M. tuberculosis SI culture-positive cases; denominator all culture-positive cases) was conducted. Diagnostic yields (95% confidence intervals, CIs) were displayed as Forest plots. Heterogeneity was evaluated using Chi-squared and I-squared tests and meta-regression analysis. Ninety publications were screened, 28 full-text papers reviewed, and 17 analyzed. Collectively, n=627 SI culture-positive cases among n=975 culture-confirmed TB cases were reported. The diagnostic yield of SI ranged from 35 to 95%. The pooled diagnostic yield was 74% (CI 65-81%), with significant heterogeneity (p<0.0001, I2=86%). There were no statistically significant differences in the yield between sub-groups defined by human immunodeficiency virus (HIV) prevalence or age. Univariate analysis demonstrated that the use of fiberoptic bronchoscopy (FOB) as the comparator method was associated with a 22% reduction (CI 2-42%) in the diagnostic yield of SI. However, after adjustment for confounding, the meta-regression analysis showed that FOB usage (p=0.21) and saline concentration (p=0.31) were not independently associated with the diagnostic yield. SI will detect approximately three-quarters of M. tuberculosis culture-positive cases under study conditions. Significant heterogeneity in the diagnostic yield was not explained by HIV prevalence, age, or the use of FOB as the comparator method. The use of a particular nebulized saline concentration for SI cannot be recommended on the basis of this meta-regression analysis.

Predictive factors for latent tuberculosis infection among adolescents in a high-burden area in South Africa

SETTING

A high tuberculosis (TB) burden area in South Africa (notification rate for all TB cases 1400 per 100 000 population).

OBJECTIVE

To determine the prevalence of and predictive factors associated with latent TB infection in adolescents.

DESIGN

Adolescents aged 12-18 years were recruited from high schools, clinical and demographic data were collected, and a tuberculin skin test (TST) and a QuantiFERON®-TB Gold In-Tube (QFT) assay performed.

RESULTS

A total of 6363 (58.2%) of 10 942 adolescents at the schools were enrolled. After exclusions, of 5244 participants, 55.2% (95%CI 53.8-56.5) had TST ≥ 5 mm, while 50.9% (49.5-52.2) were QFT-positive. On multivariate analysis, Black/mixed race racial groups, male sex, older age, household TB contact, low income and low education level were predictive factors for both TST- and QFT-positive results.

CONCLUSION

About half of the adolescents were found to be latently infected with TB in a high TB burden area with demographic and poverty-related socio-economic factors predicting the risk of TB infection. Adolescents from deprived communities should be considered an important target group for educational interventions by TB control programmes in high-burden settings.

The tuberculin skin test: a comparison of ruler and calliper readings

BACKGROUND

The Mantoux tuberculin skin test (TST) is widely used to diagnose latent infection with Mycobacterium tuberculosis. TST skin induration may be measured either by a transparent ruler or by a pair of callipers. We hypothesised that the type of instrument used may affect the reading.

OBJECTIVE

To determine whether variability in Mantoux TST measurement is affected by the type of reading instrument.

METHOD

A TST (Mantoux method) was performed among healthy adolescents. The indurations were read with among ruler and calliper by two independent readers. Limits of agreement and Kappa (κ) scores at TST positivity cut-off points were calculated. A Bland-Altman graph was constructed.

RESULTS

The 95% limits of agreement between instruments ranged from -5 mm to 3 mm. The limits of agreement between readers ranged from -5 mm to 4 mm. κ scores between instruments were respectively 0.7 and 0.8 at 15 mm and 10 mm cut-offs.

CONCLUSION

The variability between readers of TST indurations is not influenced by changing the reading instrument.

Age-related tuberculosis incidence and severity in children under 5 years of age in Cape Town, South Africa

SETTING

Limited data are available on the characteristics of tuberculosis (TB) disease in young children, especially in high-burden countries.

OBJECTIVE

To assess the incidence and severity of TB in children aged <5 years.

METHODS

TB records and chest radiographs of children born in Cape Town in 1999 and diagnosed with TB between 1999 and 2004 were reviewed retrospectively.

RESULTS

A total of 1607 cases were registered. The cumulative incidence of definite (bacteriologically confirmed) and probable (radiological evidence and > or =1 TB clinical feature) TB in children aged <5 years was 2.9% and was highest between the ages of 12 and 23 months. Of 1233 children with definite or probable TB, 506 (41%) had severe disease (dissemination, cavities or >1 lobe involved). The under 5 years incidence of disseminated TB was 0.33%. Of 239 (15%) cases that were bacteriologically confirmed, clinical features typical of TB disease were individually present in <60%. The most common combined symptoms were cough for >2 weeks and weight loss, occurring in 43/239 (18%).

CONCLUSION

TB incidence was high, and peaked in children aged 12-23 months. Many children experienced severe disease. A fifth of children with microbiologically confirmed disease presented with only one feature typically associated with TB.

Induced sputum or gastric lavage for community-based diagnosis of childhood pulmonary tuberculosis?

OBJECTIVES

To compare the diagnostic yield of Mycobacterium tuberculosis from induced sputum (IS) and gastric lavage (GL) among children in a community setting.

METHODS

Specimen-collection methods for bacteriological confirmation of pulmonary tuberculosis (PTB) were compared during a tuberculosis vaccine trial near Cape Town, South Africa (2001-2006). Children with a tuberculosis contact or compatible symptoms were investigated for suspected PTB. Diagnostic yields from 764 paired IS and GL specimens were compared in 191 culture-confirmed cases.

MEASUREMENTS AND MAIN RESULTS

The crude yield of M tuberculosis was 10.4%, n = 108 by IS (5.8%) and n = 127 by GL (6.8%), from a total of 194 cases, of which three had incomplete IS/GL specimen pairs. Agreement between IS and GL was poor (kappa = 0.31). The comparative yield of a single IS sample (38%) was equivalent to a single GL sample (42%), with a difference in yield of -4% (95% CI -15% to +7%). The combined yield of same-day IS and GL specimens (67%) was equivalent to two consecutive GL specimens (66%), with a difference in yield of 1% (95% CI -9% to 11%), but significantly greater than two consecutive IS specimens (55%), with a difference in yield of 12% (95% CI 2% to 21%). The adjusted odds of a M tuberculosis culture were increased by a positive tuberculin skin test or chest radiograph compatible with PTB.

CONCLUSIONS

In this community setting, the diagnostic yield of a single IS sample was equivalent to that of a single GL sample. The optimal diagnostic yield may be obtained from paired IS and GL specimens taken on a single day or two GL specimens taken on consecutive days.

The effect of epinephrine by nebulization on measures of airway obstruction in patients with acute severe croup

OBJECTIVES

To demonstrate that tests of pulmonary function applicable to sick infants and small children with acute severe viral croup would provide clear, objective evidence of responsiveness to therapy with nebulized epinephrine.

STUDY DESIGN

Oesophageal pressure changes and airflows at the mouth were measured in 17 patients with acute severe croup, before and after nebulization with epinephrine.

RESULTS

In 12 of the 17 patients there was a significant improvement in respiratory mechanics following epinephrine nebulization. Six of the 12 patients who responded to adrenaline also received 0.9% saline by nebulization, without improvement. No measures derived from combined flow and volume data showed any statistically significant change following epinephrine nebulization. Measures combining flow and pressure data, specifically inspiratory airway resistance, expiratory airway resistance, work of breathing, rate of work of breathing and volume for effort, showed changes of 26%, 33%, 16%, 16% and 46% respectively. The most statistically significant measures were pressure-rate product, pressure-time integral, oesophageal pressure alone and expiratory resistance. These changes persisted for at least 10 min after inhalation although there was some evidence of decline in pharmacologic effect at that time.

CONCLUSIONS

Nebulized epinephrine results in a short-lived improvement in some but not all patients with croup. This reduction in respiratory effort occurs secondary to a decline in inspiratory and expiratory airway resistance. Oesophageal pressures measured via a feeding tube are satisfactory for quantification of the acute response and may be a useful continuous monitoring device. Flow measurements are unhelpful, and continuous administration of nebulized epinephrine should be investigated.

Short-term mortality and implementation of antiretroviral treatment for critically ill HIV-infected children in a developing country

OBJECTIVE

To describe the short-term outcome of critically ill HIV-infected children with access to highly active antiretroviral therapy (HAART) in a developing region.

METHODS

Prospective observational study conducted in a paediatric teaching hospital in Cape Town, South Africa. All children admitted to the paediatric intensive care unit (PICU) with suspected HIV infection were screened. Data are n (%) with 95% confidence intervals.

RESULTS

Sixty eight of 96 HIV antibody-positive children, median age 3 months, were confirmed HIV-infected. Predicted PICU mortality was 0.42. Fifty one children (75%; 95% CI 65 to 85%) survived to PICU discharge, but hospital survival was only 51% (95% CI 40 to 63%). Limitation of intervention (LOI) decisions were a factor in the majority of PICU and ward deaths. Twenty one PICU survivors (31%; 95% CI 20 to 42%) commenced HAART, and two children were already on treatment. Nineteen children (28%) were considered to be established on HAART after 1 month. Thirteen HIV-infected children (19%; 95% CI 10 to 28%), representing 25% (95% CI 14 to 37%) of all PICU survivors, and 68% (95% CI 48 to 89%) of those PICU survivors who were established on HAART remain well on treatment after median 350 days.

CONCLUSION

The majority of HIV-infected children survived to discharge from PICU, but only half survived to hospital discharge. LOI decisions, usually made in PICU, directly influenced short-term survival and the opportunity to commence HAART. Although few critically ill HIV-infected children survived to become established on HAART, the long-term outcome of children on HAART is encouraging and warrants further investigation.

Hyperchloraemic metabolic acidosis following open cardiac surgery

AIMS

To describe acid-base derangements in children following open cardiac surgery on cardiopulmonary bypass (CPB), using the Fencl-Stewart strong ion approach.

METHODS

Prospective observational study set in the paediatric intensive care unit (PICU) of a university children's hospital. Arterial blood gas parameters, serum electrolytes, strong ion difference, strong ion gap (SIG), and partitioned base excess (BE) were measured and calculated on admission to PICU.

RESULTS

A total of 97 children, median age 57 months (range 0.03-166), median weight 14 kg (range 2.1-50), were studied. Median CPB time was 80 minutes (range 17-232). Predicted mortality was 2% and there was a single non-survivor. These children showed mild metabolic acidosis (median standard bicarbonate 20.1 mmol/l, BE -5.1 mEq/l) characterised by hyperchloraemia (median corrected Cl 113 mmol/l), and hypoalbuminaemia (median albumin 30 g/l), but no significant excess unmeasured anions or cations (median SIG 0.7 mEq/l). The major determinants of the net BE were the chloride and albumin components (chloride effect -4.8 mEq/l, albumin effect +3.4 mEq/l). Metabolic acidosis occurred in 72 children (74%) but was not associated with increased morbidity. Hyperchloraemia was a causative factor in 53 children (74%) with metabolic acidosis. Three (4%) hyperchloraemic children required adrenaline for inotropic support, compared to eight children (28%) without hyperchloraemia. Hypoalbuminaemia was associated with longer duration of inotropic support and PICU stay.

CONCLUSIONS

In these children with low mortality following open cardiac surgery, hypoalbuminaemia and hyperchloraemia were the predominant acid-base abnormalities. Hyperchloraemia was associated with reduced requirement for adrenaline therapy. It is suggested that hyperchloraemic metabolic acidosis is a benign phenomenon that should not prompt escalation of haemodynamic support. By contrast, hypoalbuminaemia, an alkalinising force, was associated with prolonged requirement for intensive care.

Evolution of an adenovirus outbreak in a multidisciplinary children's hospital

OBJECTIVE

To describe the course of an evolving adenovirus outbreak in a multidisciplinary children's hospital with a high-risk patient population.

METHODS

Observational study in a 280-bed university hospital during June 2002. Active case finding identified children with adenovirus infection. Data are median (interquartile range) or n (%). Adenovirus infection was diagnosed in 49 children, median age 12 months (4-33).

RESULTS

New cases were diagnosed over 26 days and peaked on day 17 (n = 15). Total infected inpatients peaked on days 17-21 (n = 36). Twenty-three infections (47%) were community-acquired and 26 (53%) hospital-acquired. Thirty-three children (67%) had a coexistent high-risk condition. Median hospital stay before and after diagnosis was 9 days (3-18) and 9 days (4-29), respectively. Twenty-two children (45%) were admitted to PICU. Overall hospital mortality was 22% (n = 11) and mortality attributed to adenoviral disease 12% (n = 6). Hospital mortality was similar between community- and hospital-acquired infections (22% compared to 23%) (P = 1.0). Twenty children (41%) received intravenous immunoglobulin (IVIG). Children treated with IVIG had a longer hospital stay (median 40 days vs 14 days) than those who did not receive IVIG (P = 0.01). Neither PICU mortality (29% vs 12%), nor hospital mortality (35% vs 14%), differed significantly between IVIG treated and untreated children (P = 0.76 and P = 0.16, respectively).

CONCLUSION

The rapid spread of hospital-acquired adenovirus underlines the importance of effective infection control measures. Despite nosocomial infection amongst high-risk patients, mortality was similar to that of community-acquired infection. Administration of immunoglobulin was not associated with demonstrable benefit. A prospective randomized trial would be required to resolve this issue.

HIV infection of undetermined origin during infancy

The case histories of two children with horizontally acquired HIV infection are described. These children were diagnosed at a paediatric hospital in sub-Saharan Africa. Although the source(s) of infection was not identified, both children had had several contacts with the health service, experienced invasive procedures and ingested expressed milk from their own mothers during hospital admission. Health-care institutions, particularly those located in high HIV prevalence areas, must implement effective infection control measures to ensure that the risk of horizontal infection is minimized. Attention should be given to practices that are unique to each clinical discipline.

Correction of the anion gap for albumin in order to detect occult tissue anions in shock

BACKGROUND

It is believed that hypoalbuminaemia confounds interpretation of the anion gap (AG) unless corrected for serum albumin in critically ill children with shock.

AIM

To compare the ability of the AG and the albumin corrected anion gap (CAG) to detect the presence of occult tissue anions.

METHODS

Prospective observational study in children with shock in a 22 bed multidisciplinary paediatric intensive care unit of a university childrenrsquo;s hospital. Blood was sampled at admission and at 24 hours, for acid-base parameters, serum albumin, and electrolytes. Occult tissue anions (lactate + truly "unmeasured" anions) were calculated from the strong ion gap. The anion gap ((Na + K) - (Cl + bicarbonate)) was corrected for serum albumin using the equation of Figge: AG + (0.25 x (44 - albumin)). Occult tissue anions (TA) predicted by the anion gap were calculated by (anion gap - 15 mEq/l). Optimal cut off values of anion gap were compared by means of receiver operating characteristic (ROC) curves. Ninety three sets of data from 55 children (median age 7 months, median weight 4.9 kg) were analysed. Data are expressed as mean (SD), and mean bias (limits of agreement).

RESULTS

The incidence of hypoalbuminaemia was 76% (n = 42/55). Mean serum albumin was 25 g/l (SD 8). Mean AG was 15.0 mEq/l (SD 6.1), compared to the CAG of 19.9 mEq/l (SD 6.6). Mean TA was 10.2 mmol/l (SD 6.3). The AG underestimated TA with mean bias 10.2 mmol/l (4.1-16.1), compared to the CAG, mean bias 5.3 mmol/l (0.4-10.2). A clinically significant increase of TA >5 mmol/l was present in 83% (n = 77/93) of samples, of which the AG detected 48% (n = 36/77), and the CAG 87% (n = 67/77). Post hoc ROC analysis revealed optimal cut off values for detection of TA >5 mmol/l to be AG >10 mEq/l, and CAG >15.5 mEq/l.

CONCLUSION

Hypoalbuminaemia is common in critically ill children with shock, and is associated with a low observed anion gap that may fail to detect clinically significant amounts of lactate and other occult tissue anions. We suggest that the albumin corrected anion gap should be calculated to screen for occult tissue anions in these children.

A comparison of three scoring systems for mortality risk among retrieved intensive care patients

AIMS

To assess the impact of two paediatric intensive care unit retrieval teams on the performance of three mortality risk scoring systems: pre-ICU PRISM, PIM, and PRISM II.

METHODS

A total of 928 critically ill children retrieved for intensive care from district general hospitals in the south east of England (crude mortality 7.8%) were studied.

RESULTS

Risk stratification was similar between the two retrieval teams for scores utilising data primarily prior to ICU admission (pre-ICU PRISM, PIM), despite differences in case mix. The fewer variables required for calculation of PIM resulted in complete data collection in 88% of patients, compared to pre-ICU PRISM (24%) and PRISM II (60%). Overall, all scoring systems discriminated well between survival and non-survival (area under receiver operating characteristic curve 0.83-0.87), with no differences between the two hospitals. There was a tendency towards better discrimination in all scores for children compared to infants and neonates, and a poor discrimination for respiratory disease using pre-ICU PRISM and PRISM II but not PIM. All showed suboptimal calibration, primarily as a consequence of mortality over prediction among the medium (10-30%) mortality risk bands.

CONCLUSIONS

PIM appears to offer advantages over the other two scores in terms of being less affected by the retrieval process and easier to collect. Recalibration of all scoring systems is needed.

Organophosphate poisoning complicated by a tachyarrhythmia and acute respiratory distress syndrome in a child

A 9-year-old child presented with documented organophosphate insecticide poisoning. His course was initially complicated by a tachyarrhythmia with QT-interval prolongation that responded promptly to intravenous magnesium. However, following partial recovery, he developed progressive acute respiratory distress syndrome characterized by irreversible fibrosis and obliteration of the lung parenchyma.

Presentation and outcome of severe anticholinesterase insecticide poisoning

AIMS

To document the patterns of presentation and outcome of severe anticholinesterase insecticide poisoning in children requiring intensive care.

METHODS

Retrospective case note review of all 5541 children admitted to the paediatric intensive care unit (PICU) of a university hospital during the 10 years from January 1990 to May 2000. Fifty four children (1%) with anticholinesterase insecticide poisoning were identified. Presenting features, route of exposure, treatment, complications, and mortality were recorded. Data were analysed by the Fisher's exact and Mann-Whitney tests.

RESULTS

More children than expected were from a rural area (46% versus 25%). Decontamination occurred in 50% of children prior to PICU admission. Enteral exposure was most common (n = 27; 50%). Median pseudocholinesterase level was 185 IU/l (range 75-7404). Median total dose of atropine required to maintain mydriasis was 0.3 mg/kg (range 0.03-16.7) over a median of 10 hours (range 1-160). Complications included coma (31%), seizures (30%), shock (9%), arrhythmias (9%), and respiratory failure requiring ventilation (35%). No significant differences were detected in incidence of seizures, cardiac arrhythmias, respiratory failure, mortality, duration of ventilation, or PICU stay, according to route of exposure, or state of decontamination. Four children died (7%). Mortality was associated with the presence of a cardiac arrhythmia (likelihood ratio 8.3) and respiratory failure (likelihood ratio 3.3).

CONCLUSION

The mortality and morbidity of severe anticholinesterase insecticide poisoning in childhood is not related to route of exposure, or to delay in decontamination. However, the presence of either a cardiac arrhythmia or respiratory failure is associated with a poor prognosis.

Severe upper airway obstruction caused by ulcerative laryngitis

AIMS

To present our experience of severe upper airway obstruction caused by ulcerative laryngitis in children.

METHODS

Retrospective case note review of 263 children with severe upper airway obstruction and a clinical diagnosis of croup admitted to a paediatric intensive care unit (PICU) over a five year period.

RESULTS

A total of 148 children (56%) underwent microlaryngoscopy (Storz 3.0 rigid telescope). Laryngeal ulceration with oedema was documented in 15 of these children (10%), median age 14 months (range 10-36) and median weight 10 kg (range 6-12). Twenty seven of the children who underwent microlaryngoscopy (18%) also had ulcerative gingivostomatitis consistent with herpes simplex virus infection. Ulcerative laryngitis was documented in nine of 27 (33%) children with, and in six of 121 (5%) children without, coexistent ulcerative gingivostomatitis. One of the 15 children did not require airway intervention. Nine children required nasotracheal intubation for a median of 4 days (range 3-11) and median PICU stay of 6 days (range 4-14). Five children required tracheostomy ab initio, with a median PICU stay of 30 days (range 20-36), and duration of tracheostomy in situ for a median of 19 days (range 15-253). All 15 children survived.

CONCLUSION

Ulcerative laryngitis is more common in our patient population than the few reports suggest. Early diagnostic microlaryngoscopy is recommended in children with severe croup who follow an atypical course.

Are transoesophageal Doppler parameters a reliable guide to paediatric haemodynamic status and fluid management?

OBJECTIVE

Transoesophageal Doppler (TOD) has been used in adults to optimise left ventricular filling on the basis of the waveform parameters. We wished to see if a similar relationship exists in children, specifically: (a) whether change in thermodilution stroke volume (SV) following a fluid bolus corresponded to change in Doppler stroke distance, Doppler corrected flow time (FTc), or central venous pressure (CVP); (b) whether a response to fluid challenge (defined as an increase in SV of greater than 10%) can be predicted on the basis of an absolute value for FTc or CVP prior to fluid bolus; and (c) the relationship between FTc and systemic vascular resistance index.

DESIGN

Prospective, comparison study.

SETTING

Sixteen-bed paediatric intensive care unit of a university hospital.

PATIENTS

Ninety-four ventilated children were studied, median (range) age 25 months (4 days- 16 years). Diagnoses included: post-cardiac surgery (n = 58), sepsis/multi-organ failure (n = 29), respiratory disease (n = 5), and other (n = 2).

INTERVENTIONS

A 4-MHz, 5.5-mm diameter, flexible TOD probe was placed when patients were haemodynamically stable. Five consecutive measurements of stroke distance and FTc were made and averaged, concurrently with five SV measurements by femoral artery thermodilution. SV was then augmented by administration of fluid (10 ml/kg), and haemodynamic recordings were repeated.

MEASUREMENTS AND MAIN RESULTS

The median (range) SV was 17 ml (2-64 ml). The median coefficients of variation were 3.9 % for SV, 3.5 % for stroke distance, and 3.1% for FTc. Changes in SV were accurately tracked by changes in stroke distance (mean bias 1.8 %, limits of agreement +/- 17%), but not by FTc or CVP. FTc was weakly inversely correlated with systemic vascular resistance (r = -0.15, P < 0.05). Among non-cardiac patients (n = 36), the optimal FTc that predicted an improvement in SV following fluid bolus was 0.394 s (area under ROC curve 0.756), giving a sensitivity of 90 %, specificity of 62 %, positive predictive value of 47 %, and a negative predictive value of 94 %. CVP was a poor predictor for all patient groups.

CONCLUSIONS

TOD stroke distance is able to follow changes in SV following fluid bolus amongst ventilated children, and can predict when further volume loading is unlikely to improve SV amongst general, but not cardiac ICU patients. CVP is a poor discriminator of volume status in this group of patients.

Exogenous surfactant supplementation in infants with respiratory syncytial virus bronchiolitis

Infants with respiratory syncytial virus (RSV) bronchiolitis are deficient in surfactant, both in quantity and ability to reduce surface tension. New evidence suggests surfactant has a role in maintaining the patency of conducting airways, which has implications for RSV bronchiolitis. A randomized, controlled pilot study was undertaken to assess the effects of exogenous surfactant supplementation to RSV-positive infants on pulmonary mechanics, indices of gas exchange, and the phospholipid composition of bronchoalveolar lavage fluid (BALF). Nineteen ventilated infants (median corrected age 4 wk) received either two doses of surfactant (Survanta, 100 mg/kg) within 24 and 48 h of mechanical ventilation (n = 9), or air placebo (n = 10). Static lung compliance and resistance of infants in the placebo but not in the surfactant-treated group became progressively worse over the first 30 h following enrollment. Although no significant acute changes in gas exchange parameters were seen following surfactant, infants in the surfactant group showed a more rapid improvement in oxygenation and ventilation indices over the first 60 h of ventilation. Surfactant status was assessed from the concentration ratio in BALF of the disaturated phospholipid species dipalmitoylphosphatidylcholine to that of the monounsaturated species palmitoyloleoylphosphatidylcholine. This ratio correlated with both lung compliance (positively) and resistance (negatively), and over time increased in the treated group and declined in placebo infants. The data from this pilot study suggest that functional surfactant has a role in maintaining small airway patency as well as lung compliance in infants infected with RSV and an outcome study is now warranted.

Procalcitonin and cytokine levels: relationship to organ failure and mortality in pediatric septic shock

BACKGROUND

Procalcitonin (PCT), a marker of bacterial sepsis, may also act as a mediator of the inflammatory response to infection, and thus influence outcome.

OBJECTIVE

To investigate the relationship between PCT, interleukin (IL)-10, tumor necrosis factor (TNF), organ failure, and mortality in pediatric septic shock.

DESIGN

Prospective observational study.

SETTING

A 16-bed pediatric intensive care unit of a university hospital.

PATIENTS

A total of 75 children with septic shock having a median age of 43.1 months (range, 0.1-192 months). Children who had received antibiotics for >24 hrs were excluded. A total of 37 patients (49%) had meningococcal disease, and 72 patients (96%) required mechanical ventilation.

INTERVENTIONS

The pediatric risk of mortality (PRISM) score, multiple organ system failure (MOSF) score, duration of ventilation, length of ICU stay, and outcome were recorded. PCT, IL-10, and TNF were measured at admission to the intensive care unit. Sequential PCT levels were available at 0 hrs and 24 hrs in 39 patients (52%).

RESULTS

Observed mortality was 21/75 (28%). Data are median (range). The admission PCT (p = .0002) and TNF levels (p = .0001) were higher in children with higher MOSF scores. In survivors and nonsurvivors, the admission PCT was 82 ng/mL vs. 273 ng/mL (p = .03), IL-10 was 62 pg/mL vs. 534 pg/mL (p = .03), and TNF was 76 pg/mL vs. 480 pg/mL (p = .001), respectively. Area under the mortality receiver operating characteristic curve was 0.73 for PCT, 0.67 for IL-10, and 0.76 for TNF, compared with 0.83 for the PRISM score. Of 39 children, 16 (41%) with sequential PCT measurements showed no fall in PCT after 24 hrs treatment. These children had higher admission levels of IL-10 (p = .03), and TNF (p = .03) compared with children who demonstrated a subsequent fall in PCT. Although the former did not have a higher median PRISM (p = .28) or MOSF score (p = .19), observed mortality was 44% (7 of 16) compared with 9% (2 of 23) (p = .02).

CONCLUSION

The admission PCT, like TNF and IL-10, is related to the severity of organ failure and mortality in children with septic shock. A fall in PCT after 24 hrs of treatment may have favorable prognostic significance.

Evaluation of the 5-French saline paediatric gastric tonometer

OBJECTIVE

To evaluate the paediatric 5-French (Fr) saline-filled gastric tonometer.

DESIGN

(a) In vitro comparison of saline bath reference pCO2 with tonometric pCO2 measured by normal saline-filled and phosphate-buffered saline-filled 5-Fr tonometers, and by a recirculating gas tonometer. ( b) In vivo comparison of gastric intramucosal pCO2i, measured by normal saline-filled 5-Fr tonometer (NST) and simultaneously by recirculating gas tonometer (RGT) in ten paediatric intensive care patients. (c) In vivo comparison of pCO2i measured simultaneously by 2 NST 5-Fr tonometers, before and after enteral feeding, in ten paediatric intensive care patients.

MEASUREMENTS AND MAIN RESULTS

(a) Twenty consecutive measurements of pCO2 were made at constant reference pCO2 of 19, 38, 56, and 75 mmHg (2.5, 5.0, 7.5, and 10.0 kPa), respectively. The NST tonometer underestimated reference pCO2 by mean bias (limits of agreement) of 58% (20%), and the phosphate-buffered saline-filled tonometer by 6% (26%). The RGT showed mean bias 5.7% with narrow limits of agreement (1.5%). (b) In 50 paired (NST vs. RGT) in vivo measurements over pCO2i range 23-73 mmHg (3.0-9.7 kPa), the NST underestimated RGT pCO2i by a mean bias of 10 mmHg (1.3 kPa), with limits of agreement +/-10 mmHg (1.5 kPa). This resulted in NST consistently overestimating pHi and underestimating pCO2 gap (both P < 0.001). (c) One hundred simultaneous paired NST measurements were assessed (50 without, and 50 with enteral feeding). The mean biases (limits of agreement) were identical in the fasted and fed states 0.4+/-6 mmHg, with no difference between the fed and fasting states (P = 0.7).

CONCLUSIONS

There are inherent problems in the methodology of saline tonometry, which adversely affect the accuracy and reliability of the 5-Fr paediatric gastric tonometer in comparison to recirculating gas tonometry.

Use of transesophageal Doppler ultrasonography in ventilated pediatric patients: derivation of cardiac output

OBJECTIVE

To ascertain if cardiac output (CO) could be derived from blood flow velocity measured in the descending aorta of ventilated children by transesophageal Doppler ultrasonography (TED) without the need for direct aortic cross sectional area measurement, and to evaluate the ability of TED to follow changes in CO when compared with femoral artery thermodilution.

DESIGN

Prospective, comparison study.

SETTING

A 16-bed pediatric intensive care unit of a university hospital.

PATIENTS

A total of 100 ventilated infants and children aged 4 days to 18 yrs (median age, 27 months). Diagnoses included postcardiac surgery (n = 58), sepsis/multiple organ failure (n = 32), respiratory disease (n = 5), and other (n = 5). A total of 55 patients were receiving inotropes or vasodilators.

INTERVENTIONS

When patients were hemodynamically stable, a TED probe was placed into the distal esophagus to obtain optimal signal, and minute distance (MD) was recorded. Five consecutive MD measurements were made concurrently with five femoral artery thermodilution measurements, and the concurrent measurements were averaged. CO was then manipulated by fluid administration or inotrope adjustment, and the readings were repeated.

MEASUREMENTS AND MAIN RESULTS

Femoral artery thermodilution CO ranged from 0.32 to 9.19 L/min, (median, 2.46 L/min), and encompassed a wide range of high and low flow states. Theoretical consideration revealed the optimal TED estimate for CO to be (MD x patient height2 x 10(-7)). Linear regression analysis yielded a power function model such that: estimated CO = 1.158 x (MD x height2 x 10(-7))(0.785), r2 = 0.879, standard error of the estimate = 0.266. Inclusion of a correction factor for potential changes in aortic cross-sectional area with hypo- and hypertension did not appreciably improve the predictive value of the model. MD was able to follow percentage changes in CO, giving a mean bias of 0.87% (95% confidence interval -0.85% to 2.59%), and limits of agreement of +/- 16.82%. The median coefficient of variation for MD was 3.3%.

CONCLUSIONS

TED provides a clinically accurate estimate of CO across the entire pediatric age range and is able to follow changes in CO.

Early hyperlactataemia in critically ill children

OBJECTIVE

To examine the relationships between early hyperlactataemia, acidosis, organ failure, and mortality in children admitted to intensive care.

DESIGN

Prospective observational study. Children with lactate levels > 2 mmol/l were eligible for enrolment. Post-operative patients and those with inherited metabolic disease were excluded. Seven hundred and five children admitted to intensive care were screened, and 50 children with hyperlactataemia (incidence 7%), aged 20.3 months (0.1-191) were enrolled and followed up. The Paediatric Risk of Mortality (PRISM) score, Multiorgan System Failure (MOSF) score, length of ICU stay, and outcome were recorded. Data were collected for lactate (mmol/l), pH, and base excess (BE) until 24 h after admission. Data are reported as median (range) and were analysed by the Mann-Whitney, Fisher's Exact, and Kruskal-Wallis tests, and chisquared test for trend.

RESULTS

Overall mortality in the screening group was 70/705 (10%). In the study group (n = 50) median PRISM score was 19 (4-49), median MOSF score 2 (1-4), and observed mortality 32/50 (64%). Median duration of ICU stay was 6 days (2-32) in survivors, and median time until death 3 days (0-13) in nonsurvivors. Eleven nonsurvivors (34%) died within 24 h. In the screening group, hyperlactataemia on admission identified mortality with likelihood ratio = 15. In the study group, neither the admission lactate (3.8 vs 4.6 mmol/l, P = 0.27), pH (7.32 vs 7.30, P = 0.6), nor BE (-7.5 vs -8, P = 0.45) differed significantly between survivors and nonsurvivors. Neither the admission nor peak lactate increased with increasing MOSF score (P = 0.5 and 0.54). The median peak lactate level was 5 mmol/l (2-9.3) in survivors compared to 6.8 mmol/l (2.3-22) in nonsurvivors (P = 0.02), and the cumulative average lactate level was 2.4 mmol/l (1-4.9) in survivors, compared to 4.5 mmol/l (1.6-21) in nonsurvivors (P = 0.0003). Persistent hyperlactataemia 24 h after admission identified mortality with likelihood ratio = 7.

CONCLUSION

Hyperlactataemia on admission to intensive care is associated with a high mortality in children. Nonsurvivors within this group may be distinguished by the peak lactate level, or by persistent hyperlactataemia after 24 h of treatment.

Procalcitonin may help differentiate disseminated herpes simplex viral infection from bacterial sepsis in neonates

Disseminated herpes simplex virus infection is a potentially fatal condition which may be difficult to differentiate from bacterial sepsis. We report the case of a neonate with overwhelming herpes simplex (type 2) viraemia who presented with 'septic shock'.

CONCLUSION

A low procalcitonin level (1.6 ng/ml), inconsistent with bacteraemia, suggests an alternative aetiology and may strengthen the case for antiviral therapy.

Diagnostic markers of infection: comparison of procalcitonin with C reactive protein and leucocyte count

BACKGROUND

Procalcitonin has been advocated as a marker of bacterial infection.

OBJECTIVE

To evaluate diagnostic markers of infection in critically ill children, comparing procalcitonin with C reactive protein and leucocyte count in a paediatric intensive care unit (PICU).

METHODS

Procalcitonin, C reactive protein, and leucocyte count were measured in 175 children, median age 16 months, on admission to the PICU. Patients were classified as: non-infected controls (43); viral infection (14); localised bacterial infection without shock (25); bacterial meningitis/encephalitis (10); or septic shock (77). Six children with "presumed septic shock" (without sufficient evidence of infection) were analysed separately. Optimum sensitivity, specificity, predictive values, and area under the receiver operating characteristic (ROC) curve were evaluated.

RESULTS

Admission procalcitonin was significantly higher in children with septic shock (median 94.6; range 3.3-759.8 ng/ml), compared with localised bacterial infection (2.9; 0-24.3 ng/ml), viral infection (0.8; 0-4.4 ng/ml), and non-infected controls (0; 0-4.9 ng/ml). Children with bacterial meningitis had a median procalcitonin of 25.5 (7.2-118.4 ng/ml). Area under the ROC curve was 0.96 for procalcitonin, 0.83 for C reactive protein, and 0.51 for leucocyte count. Cut off concentrations for optimum prediction of septic shock were: procalcitonin > 20 ng/ml and C reactive protein > 50 mg/litre. A procalcitonin concentration > 2 ng/ml identified all patients with bacterial meningitis or septic shock.

CONCLUSION

In critically ill children the admission procalcitonin concentration is a better diagnostic marker of infection than C reactive protein or leucocyte count. A procalcitonin concentration of 2 ng/ml might be useful in differentiating severe bacterial disease in infants and children.