Data Science Solution to Event Prediction in Outsourced Clinical Trial Models

Late phase clinical trials are regularly outsourced to a Contract Research Organisation (CRO) while the risk and accountability remain within the sponsor company. Many statistical tasks are delivered by the CRO and later revalidated by the sponsor. Here, we report a technological approach to standardised event prediction. We have built a dynamic web application around an R-package with the aim of delivering reliable event predictions, simplifying communication and increasing trust between the CRO and the in-house statisticians via transparency. Short learning curve, interactivity, reproducibility and data diagnostics are key here. The current implementation is motivated by time-to-event prediction in oncology. We demonstrate a clear benefit of standardisation for both parties. The tool can be used for exploration, communication, sensitivity analysis and generating standard reports. At this point we wish to present this tool and share some of the insights we have gained during the development.

Clinical evaluation of the efficacy of the P2X7 purinergic receptor antagonist AZD9056 on the signs and symptoms of rheumatoid arthritis in patients with active disease despite treatment with methotrexate or sulphasalazine

OBJECTIVES

The P2X(7) purinergic receptor antagonist AZD9056 was evaluated in a phase IIa study and subsequently in a phase IIb study to assess the effects of orally administered AZD9056 on the signs/symptoms of rheumatoid arthritis (RA), with American College of Rheumatology 20% response criteria (ACR20) as the primary outcome.

METHODS

Both studies were randomised, double-blind, placebo-controlled, parallel-group studies in patients with RA receiving methotrexate or sulphasalazine. Phase IIa was an ascending-dose trial in two cohorts (n=75) using AZD9056 administered daily over 4 weeks. Phase IIb included an open-label etanercept treatment group. Patients were randomised to receive treatment for 6 months with 50, 100, 200 or 400 mg AZD9056 (oral, once a day) or matching placebo (oral, once a day), or subcutaneous etanercept (50 mg once a week).

RESULTS

In phase IIa, 65% of AZD9056 recipients at 400 mg/day responded at the ACR20 level compared with 27% of placebo-treated patients. A significant reduction in swollen and tender joint count was observed in the actively treated group compared with placebo, whereas no effect on acute-phase response was observed. Of 385 randomised patients in the phase IIb study, 383 received treatment. AZD9056 (all doses) had no clinically or statistically significant effect on RA relative to placebo as measured by the proportion of patients meeting the ACR20 criteria at 6 months and further supported by secondary end points. In both studies AZD9056 was well tolerated up to 400 mg/day.

CONCLUSIONS

AZD9056 does not have significant efficacy in the treatment of RA, and the P2X(7) receptor does not appear to be a therapeutically useful target in RA.

TRIAL REGISTRATION NUMBER

ClinicalTrials.gov NCT00520572.

International albumin use: 1995 to 2006

Over the last ten years more reliable information regarding the risks and benefits of the use of albumin for fluid resuscitation has emerged. To determine what influence this has had on clinical practice, we sought to document albumin use (from mass of albumin supplied to hospitals) in 16 industrialised countries between 1995 and 2006. Data on national albumin and synthetic colloid use was sought from independent intensive care researchers and albumin issuers. The mass of albumin supplied per 10,000 persons on an annual basis by country and aggregated across the study countries was calculated. Volumes of synthetic colloid supplied per 10,000 persons were calculated. Data were obtained for 15 countries. Albumin use varied significantly between countries and throughout the observation period. Overall, aggregate albumin use decreased from a peak of 2.54 kg per 10,000 persons in 1995 to 1.40 kg per 10,000 persons in 1999; use has remained relatively constant since. Data on supply of synthetic colloids was available in only three countries and varied from 11.7 litres per 10,000 persons in Canada in 1995, to 231.8 litres per 10,000 persons in Denmark in 2004. Between 1995 and 1999 albumin use decreased and has been materially constant since; where data were available, use of synthetic colloids increased. Whether these practice changes have resulted in a net health gain or in harm requires further research.

Surgical excision margins for primary cutaneous melanoma

BACKGROUND

Cutaneous melanoma accounts for 75% of skin cancer deaths. Standard treatment is surgical excision with a safety margin some distance from the borders of the primary tumour. The purpose of the safety margin is to remove both the complete primary tumour and any melanoma cells that might have spread into the surrounding skin.Excision margins are important because there could be trade-off between a better cosmetic result but poorer long-term survival if margins become too narrow. The optimal width of excision margins remains unclear. This uncertainty warrants systematic review.

OBJECTIVES

To assess the effects of different excision margins for primary cutaneous melanoma.

SEARCH STRATEGY

In August 2009 we searched for relevant randomised trials in the Cochrane Skin Group Specialised Register; the Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library (Issue 3, 2009), MEDLINE, EMBASE, LILACS, and other databases including Ongoing Trials Registers.

SELECTION CRITERIA

We considered all randomised controlled trials (RCTs) of surgical excision of melanoma comparing different width excision margins.

DATA COLLECTION AND ANALYSIS

We assessed trial quality, and extracted and analysed data on survival and recurrence. We collected adverse effects information from included trials.

MAIN RESULTS

We identified five trials. There were 1633 participants in the narrow excision margin group and 1664 in the wide excision margin group. Narrow margin definition ranged from 1 to 2 cm; wide margins ranged from 3 to 5 cm. Median follow-up ranged from 5 to 16 years.

AUTHORS' CONCLUSIONS

This systematic review summarises the evidence regarding width of excision margins for primary cutaneous melanoma. None of the five published trials, nor our meta-analysis, showed a statistically significant difference in overall survival between narrow or wide excision.The summary estimate for overall survival favoured wide excision by a small degree [Hazard Ratio 1.04; 95% confidence interval 0.95 to 1.15; P = 0.40], but the result was not significantly different. This result is compatible with both a 5% relative reduction in overall mortality favouring narrower excision and a 15% relative reduction in overall mortality favouring wider excision. Therefore, a small (but potentially important) difference in overall survival between wide and narrow excision margins cannot be confidently ruled out.The summary estimate for recurrence free survival favoured wide excision [Hazard Ratio 1.13; P = 0.06; 95% confidence interval 0.99 to 1.28] but again the result did not reach statistical significance (P < 0.05 level).Current randomised trial evidence is insufficient to address optimal excision margins for primary cutaneous melanoma.

Anticholinergics for urinary symptoms in multiple sclerosis

BACKGROUND

Multiple Sclerosis (MS) is the commonest physically disabling chronic neurological disease affecting young people. Urinary symptoms are present in about 68% of people with MS but their basis has a number of potential aetiologies that can change with time.

OBJECTIVES

To assess the absolute and comparative efficacy, tolerability and safety of anticholinergic agents in MS patients.

SEARCH STRATEGY

We searched the Cochrane Multiple Sclerosis Group Specialised Trials Register, the Cochrane Central Register of Controlled Trials (The Cochrane Library 2008, Issue1), MEDLINE (January 1966 to January 2008), EMBASE (January 1974 to January 2008), supplemented with search of reference lists, personal communication with authors and relevant drug manufacturers.

SELECTION CRITERIA

Randomised trials and cross-over trials (blinded and unblinded) that are either placebo-controlled or comparing two or more treatments.

DATA COLLECTION AND ANALYSIS

All four review authors independently assessed eligibility and trial quality, and extracted data. Data were processed as described in the Cochrane Handbook for Systematic Reviews of Interventions.

MAIN RESULTS

Our search strategy identified 33 articles of which thirty were excluded. Three single centre trials were included. No details were given regarding randomisation and blinding in the first two trials but side effects were frequent with all treatments.The first (Hebjorn 1977) was a double blind randomised crossover trial. Thirty four persons with MS received three drugs Methantheline Bromide, Flavoxate Chloride and Meladrazine Tartrate each for 14 days, washout periods were not mentioned. Median volume measurements at the first bladder contraction were statistically significant at a 5% level for Methantheline Bromide only compared to no treatment.The second (Gajewski 1986) was a prospective parallel group randomised study. Thirty four persons with MS were treated for 6-8 weeks with Oxybutynin (19 subjects) or Propantheline (15 subjects). For frequency, nocturia, urgency, and urge incontinence differences in symptom grade in favour of Oxybutynin were found. However, only for frequency the difference was statistically significant at 5% level.The third (Fader 2007) was a double blind crossover trial. Sixty four persons with MS received oral Oxybutynin or intravesical Atropine for 14 days. Details of randomisation and blinding were given. There was no significant difference between the two treatments in any efficacy outcome measure. Side effects and QOL scores showed significant differences in favour of atropine.

AUTHORS' CONCLUSIONS

From the available evidence we cannot advocate the use of anticholinergics in MS.

Topical treatments for fungal infections of the skin and nails of the foot

BACKGROUND

Fungal infections of the feet normally occur in the outermost layer of the skin (epidermis). The skin between the toes is a frequent site of infection which can cause pain and itchiness. Fungal infections of the nail (onychomycosis) can affect the entire nail plate.

OBJECTIVES

To assess the effects of topical treatments in successfully treating (rate of treatment failure) fungal infections of the skin of the feet and toenails and in preventing recurrence.

SEARCH STRATEGY

We searched the Cochrane Skin Group Specialised Register (January 2005), the Cochrane Central Register of Controlled Trials (The Cochrane Library Issue 1, 2005), MEDLINE and EMBASE (from inception to January 2005). We screened the Science Citation Index, BIOSIS, CAB - Health and Healthstar, CINAHL DARE, NHS Economic Evaluation Database and EconLit (March 2005). Bibliographies were searched.

SELECTION CRITERIA

Randomised controlled trials (RCTs) using participants who had mycologically diagnosed fungal infections of the skin and nails of the foot.

DATA COLLECTION AND ANALYSIS

Two authors independently summarised the included trials and appraised their quality of reporting using a structured data extraction tool.

MAIN RESULTS

Of the 144 identified papers, 67 trials met the inclusion criteria. Placebo-controlled trials yielded the following pooled risk ratios (RR) of treatment failure for skin infections: allylamines RR 0.33 (95% CI 0.24 to 0.44); azoles RR 0.30 (95% CI 0.20 to 0.45); ciclopiroxolamine RR 0.27 (95% CI 0.11 to 0.66); tolnaftate RR 0.19 (95% CI 0.08 to 0.44); butenafine RR 0.33 (95% CI 0.24 to 0.45); undecanoates RR 0.29 (95% CI 0.12 - 0.70). Meta-analysis of 11 trials comparing allylamines and azoles showed a risk ratio of treatment failure RR 0.63 (95% CI 0.42 to 0.94) in favour of allylamines. Evidence for the management of topical treatments for infections of the toenails is sparser. There is some evidence that ciclopiroxolamine and butenafine are both effective but they both need to be applied daily for prolonged periods (at least 1 year). The 6 trials of nail infections provided evidence that topical ciclopiroxolamine has poor cure rates and that amorolfine might be substantially more effective but more research is required.

AUTHORS' CONCLUSIONS

Placebo-controlled trials of allylamines and azoles for athlete's foot consistently produce much higher percentages of cure than placebo. Allylamines cure slightly more infections than azoles and are now available OTC. Further research into the effectiveness of antifungal agents for nail infections is required.

Systematic review of rosacea treatments

BACKGROUND

Rosacea is a common chronic skin and ocular condition. It is unclear which treatments are most effective. We have conducted a Cochrane review of rosacea therapies. This article is a distillation of that work.

OBJECTIVE

We sought to assess the evidence for the efficacy and safety of rosacea therapies.

METHODS

Multiple databases were systematically searched. Randomized controlled trials in people with moderate to severe rosacea were included. Study selection, assessment of methodologic quality, data extraction, and analysis were carried out by two independent researchers.

RESULTS

In all, 29 studies met inclusion criteria. Topical metronidazole is more effective than placebo (odds ratio 5.96, 95% confidence interval 2.95-12.06). Azelaic acid is more effective than placebo (odds ratio 2.45, 95% confidence interval 1.82-3.28). Firm conclusions could not be drawn about other therapies.

LIMITATIONS

The quality of the studies was generally poor.

CONCLUSIONS

There is evidence that topical metronidazole and azelaic acid are effective. There is some evidence that oral metronidazole and tetracycline are effective. More well-designed, randomized controlled trials are required to provide better evidence of the efficacy and safety of other rosacea therapies.

The effect of pre-existing medical conditions and age on mortality after injury

BACKGROUND

Pre-existing medical conditions (PMCs) have been shown to increase mortality after trauma even after adjustment for the effect of chronological aging. It has been suggested that there is an interaction between injury severity and physiologic reserve, such that diminished physiologic reserve will have an adverse effect on survival at lower injury severity, but that at higher levels of injury severity, physiologic reserve will have much less of an impact.

METHODS

Records of 65,743 patients, admitted after trauma, were extracted from the database of the United Kingdom Trauma Network to explore the impacts of age, gender and PMCs on mortality, and modification of these effects by severity of injury.

RESULTS

PMCs were categorized as absent (23%), present (23%), or unrecorded (54%). There was an increase in mortality with increasing age at all levels of injury severity. Presence of a PMC was associated with a marked increase in mortality of patients with minor injuries (odds ratio [OR] = 5.9, 95% confidence interval [CI] 4.4, 8.0) or moderate injuries (OR = 2.0, 95% CI 1.4, 2.9), but not in those with more severe injuries (OR = 1.1, 95% CI 0.9, 1.4). The impact of age and male gender were also somewhat more pronounced for patients with less severe injuries.

CONCLUSION

These findings support the hypothesis of an interaction between physiologic reserve and injury severity, where PMCs are associated with increased mortality when combined with low to moderate severity injuries, but not when combined with more severe injuries.

Outcome prediction in trauma

BACKGROUND

In the Trauma Audit and Research Network (TARN), currently the largest trauma network in Europe, outcome prediction is performed using the TRISS methodology since 1989. Its database contains 200,000 hospital admissions from 110 hospitals over the country, but a large amount of data is lost for the modelling because of missing data. To improve some of the shortcomings of TRISS a new model was developed.

METHODS

The data for modelling consisted of 100,399 hospital trauma admissions over the period 1996 to 2001. Using the Glasgow Coma Score (GCS) instead of RTS has dramatically reduced the number of missing cases. Gender and its interaction with age have also been included in the model. The model was tested on different subsets of cases traditionally excluded, such as children, those with penetrating injuries, and ventilated and transferred patients. The new model included all those subsets using age, a transformation of ISS, GCS, gender and gender by age interaction as predictors.

RESULTS

The model has shown a good discriminant ability tested by the area under the receiver operating characteristic (AROC) curve. The values of the AROC for the new model were 0.947 (95% CI: 0.943-0.951) on the prediction set and 0.952 (95% CI: 0.946-0.957) on the validation set compared respectively with 0.937 (95% CI: 0.932-0.943) and 0.941 (95% CI: 0.936-0.952) for TRISS.

CONCLUSION

The new model has enabled us to include most of the cases that were excluded under the TRISS's inclusion criteria, less missing data are incurred and the predictive performance was significantly better than that of the TRISS model as shown by the AROC curves.

A New Approach to Outcome Prediction in Trauma: A Comparison With the TRISS Model

BACKGROUND

The Trauma Audit & Research Network (TARN) has been using the TRISS methodology since 1989. Its database contains 200,000 hospital admissions from 110 hospitals over the country. To improve outcome prediction, a revision of the current model became necessary. Our model tried to overcome some of the concerns of the trauma community, namely missing data, functional form of the predictors, inclusion criteria and patient's death within 30 days.

METHODS

The data for modeling consisted of 100,399 anonymized hospital trauma admissions during the period 1996 to 2001. Cross validation was performed on this data set, and a multiple logistic regression model was derived using the prediction set and then its prediction ability was tested on the validation set. Fractional polynomials modeling showed that the linear functional form of the Injury Severity Score (ISS) in the model was not satisfactory. Using the Glasgow Coma Score (GCS) instead of the revised trauma score (RTS) has dramatically reduced the number of missing cases. Sex and its interaction with age have also been included in the model. The model was tested on different subsets of cases, traditionally excluded, such as children, those with penetrating injuries, and ventilated and transferred patients. The new model included all those subsets using age, a transformation of ISS, GCS, sex, and sex by age interaction as predictors.

RESULTS

The model has shown a good discriminant ability tested by the Area under the Receiver Operating Characteristic (AROC) curve. The values of the AROC for the new model were 0.947 (95% confidence interval [CI]: 0.943-0.951) on the prediction set and 0.952 (95% CI: 0.946-0.957) on the validation set compared, respectively, with 0.937 (95% CI: 0.932-0.943) and 0.941 (95% CI: 0.936-0.952) for TRISS.

CONCLUSION

The new model has enabled us to include most of the cases that were excluded under the TRISSs inclusion criteria, less missing data are incurred and the predictive performance was significantly better than that of the TRISS model as shown by the AROC curves.

A randomized controlled trial comparing the accuracy of general diagnostic upper gastrointestinal endoscopy performed by nurse or medical endoscopists

BACKGROUND AND STUDY AIMS

Rising demand for general diagnostic upper gastrointestinal endoscopy in the UK is outgrowing the capacity of doctors to provide this service within a reasonable time. One solution is to train nurses to carry out the procedure, but it is not known whether nurses can perform general diagnostic upper gastrointestinal endoscopy as competently as doctors.

PATIENTS AND METHODS

A randomized controlled non-inferiority trial compared the adequacy and the accuracy of diagnostic upper gastrointestinal endoscopies performed by five medical and two nurse endoscopists. The videotaped procedures were assessed by a consultant gastroenterologist blinded to the identity of the endoscopist.

RESULTS

641 patients were randomly allocated (before attendance and consent procedure) to endoscopy carried out either by a doctor or a nurse. Of these, 412 were enrolled and 367 (89 %) were included in the analysis. An adequate view was obtained throughout in 53.4 % (93/177) of doctor endoscopies and 91.6 % (174/190) of nurse endoscopies (difference 38.2 %, 95 % CL 30.5 %, 47.2 %). In adequately viewed areas, the mean agreement between doctor and expert was 81.0 % and between nurse and expert it was 78.3 % (difference between the means 2.7 %, 95 % CL - 1.0 %, 6.4 %). There was no difference between doctors and nurses in the rate of biopsy performance (90.4 % and 91.1 %, respectively, P = 0.862). Nurses took longer (8.1 minutes vs. 4.6 minutes, P < 0.001) and used intravenous sedation more often (57.6 %, P = 0.027). Adequacy of view correlated positively with endoscopy duration ( P < 0.001), but diagnostic accuracy correlated inversely with duration ( P < 0.001). Neither adequacy or accuracy correlated significantly with use of intravenous sedation.

CONCLUSIONS

In endoscopies performed by nurses, the proportion of adequate examinations was much higher than that found for doctors. In areas with an adequate view, there is no significant difference in accuracy between nurses and doctors. Nurses can provide an accurate general diagnostic upper gastrointestinal endoscopy service as competently as doctors.

Identification, Preparation, and Characterization of Several Polymorphs and Solvates of Terazosin Hydrochloride

The phenomenon of polymorphism is prevalent in pharmaceuticals, yet it is unusual to identify more than three or four forms for any particular drug. Terazosin hydrochloride has been found to exist at room temperature in four solvent-free forms that can be isolated directly, one solvent-free form that can be prepared by desolvation of a methanolate, a methanol solvate, and a dihydrate. This study presents characterization and methods for preparation of each of these forms. Data are also presented demonstrating the relative stability of these forms.

Interventions for chronic palmoplantar pustulosis

BACKGROUND

Chronic palmoplantar pustulosis (PPP) is a chronic inflammatory skin condition characterised by crops of sterile pustules (yellow pus spots) on the palms and soles which erupt repeatedly over months or years. The affected areas tend to become red and scaly; cracks may form and these are often painful. Many different treatments have been used for palmoplantar pustulosis but none is generally accepted as being reliably effective.

OBJECTIVES

To assess the effects of treatments for palmoplantar pustulosis, both in reducing disease severity and in maintaining remission once achieved.

SEARCH STRATEGY

We searched the Cochrane Skin Group Specialised Register (January 2003), the Cochrane Central Register of Controlled Trials (The Cochrane Library Issue 1, 2003), MEDLINE (1966 to February 2003), EMBASE (1988 to February 2003). We also cross-checked with the Salford Database of Psoriasis Trials and reference lists of articles. We also contacted authors included trials, members of the Cochrane Skin Group and dermatologists interested in psoriasis.

SELECTION CRITERIA

Any randomised controlled trial in which patients with chronic palmoplantar pustulosis were randomised to receive one or more interventions.

DATA COLLECTION AND ANALYSIS

At least two reviewers independently assessed trial eligibility and quality. Study authors were contacted for additional information. Adverse effects information was collected from the trials.

MAIN RESULTS

Twenty-three trials involving 724 people were included. There is evidence supporting the use of systemic retinoids (improvement rate difference 44%, 95 CI 28 to 59%), oral PUVA (improvement rate difference 44%, 95 CI 26 to 62%). However, a combination of PUVA and retinoids is better than the individual treatments. The use of topical steroid under hydrocolloid occlusion is beneficial. It would also appear that low dose ciclosporin, tetracycline antibiotics and Grenz Ray Therapy may be useful in treating PPP. Colchicine has a lot of side effects and it is unclear if it is effective and neither was topical PUVA (rate difference of 0.00, 95% CI -0.04 to +0.04). There is no evidence to suggest that short-term treatment with hydroxycarbamide (hydroxyurea) is effective.

AUTHORS' CONCLUSIONS

Many different interventions were reported to produce "improvement" in PPP. There is, however, no standardised method for assessing response to treatment, and reductions in pustule counts or other empirical semi-quantitative scoring systems may be of little relevance to the patient. This review has shown that the ideal treatment for PPP remains elusive and that the standards of study design and reporting need to be improved to inform patients and those treating them of the relative merits of the many treatments available to them.

Interventions for bullous pemphigoid

BACKGROUND

Bullous pemphigoid is the most common autoimmune bullous disease in the West. Oral steroids are considered the standard treatment.

OBJECTIVES

To assess the effects of treatments for bullous pemphigoid.

SEARCH STRATEGY

We searched the Skin Group Specialised Register, the Cochrane Central Register of Controlled Trials, MEDLINE, EMBASE to March 2003 and bibliographies from identified studies.

SELECTION CRITERIA

Randomised controlled trials of treatments for patients with immunofluorescence confirmed bullous pemphigoid.

DATA COLLECTION AND ANALYSIS

Two reviewers evaluated the studies in terms of the inclusion criteria, five extracted data independently; disagreements were resolved by discussion. Statistical pooling of the data was inappropriate because of heterogeneity of treatments.

MAIN RESULTS

We found seven randomised controlled trials with a total of 634 patients. All studies involved different comparisons, none included a placebo group. Different doses, different formulations of corticosteroids and the addition of azathioprine failed to show significant differences in measures of disease control. However, patients who took azathioprine were able to almost halve the amount of prednisone required for disease control. Plasma exchange plus prednisone achieved significantly better disease control than prednisone alone; this favourable effect was not apparent in another study. The latter study also compared plasma exchange or azathioprine plus prednisone, but failed to show significant differences for disease control or mortality, although total adverse events at six months almost reached statistical significance in favour of plasma exchange plus prednisone. Comparing tetracycline plus nicotinamide with prednisolone, no significant difference for disease response was shown. A very potent topical corticosteroid was compared to oral prednisone in patients with moderate and extensive disease. In patients with extensive disease, the topical steroid group showed significantly better survival and disease control, and less severe complications, while no significant differences for these outcomes were seen in patients with moderate disease. Most of the reported deaths were in patients taking high doses of oral corticosteroids.

AUTHORS' CONCLUSIONS

Very potent topical steroids are effective and safe treatments for bullous pemphigoid; their use in extensive disease may be limited by side effects and practical factors. Starting doses of prednisolone greater than 0.75 mg/kg/day do not seem to give additional benefit, lower doses may be adequate for disease control; this could reduce the incidence and severity of adverse reactions. The effectiveness of the addition of plasma exchange or azathioprine to corticosteroids has not been established. Combination treatment with tetracycline and nicotinamide may be useful; this needs further validation.

Interventions for rosacea

BACKGROUND

Rosacea is a common chronic skin condition affecting the face, characterised by flushing, redness, pimples, pustules, and dilated blood vessels. The eyes are often involved. Frequently it can be controlled, but it is not clear which treatments are most effective.

OBJECTIVES

To assess the evidence for the efficacy and safety of treatments for rosacea.

SEARCH STRATEGY

We searched the Skin Group Specialised Register (February 2005), Cochrane Central Register of Controlled Trials (The Cochrane Library Issue 1, 2005), MEDLINE (1966 to February 2005), EMBASE (1980 to February 2005), BIOSIS (1970 to March 2002) and the Science Citation Index (1988 to February 2005). Reference lists of trials and key review articles were searched. Relevant manufacturers and experts were contacted.

SELECTION CRITERIA

Randomised controlled trials in people with moderate to severe rosacea were included. Studies judged by the authors to have seriously flawed methodology were excluded.

DATA COLLECTION AND ANALYSIS

Study selection, assessment of methodological quality, data extraction and analysis were carried out by two independent authors. Disagreements were resolved by discussion and consensus.

MAIN RESULTS

The evidence provided by twenty-nine included studies was generally weak because of poor methodology and reporting. One of our primary outcome measures, 'quality of life', was not assessed in any of the studies. Only two studies of ocular rosacea were included. Pooled data from two trials involving 174 participants indicated that according to the participants, topical metronidazole is more effective than placebo (odds ratio (OR) 5.96, 95% confidence interval (CI) 2.95 to 12.06). Data pooled from three between-patient trials showed a clear improvement in the azelaic acid group; the rates of treatment success were approximately 70 to 80% versus 50% to 55% (OR 2.45, 95% CI 1.82 to 3.28). A within-patient trial of azelaic cream versus placebo could not be pooled with the other three studies, but also showed good evidence of efficacy. Data pooled from three studies of oral tetracycline versus placebo involving 152 participants showed that, according to physicians, tetracycline was effective (OR 6.06, 95% CI 2.96 to 12.42). Some evidence of efficacy of oral metronidazole was provided by one small study.

AUTHORS' CONCLUSIONS

The quality of studies evaluating rosacea treatments was generally poor. There is evidence that topical metronidazole and azelaic acid are effective. There is some evidence that oral metronidazole and tetracycline are effective. There is insufficient evidence concerning the effectiveness of other treatments. Good RCTs looking at these treatments are urgently needed.

Uncertainty method improved on best–worst case analysis in a binary meta-analysis

BACKGROUND

Most systematic reviewers aim to perform an intention-to-treat meta-analysis, including all randomized participants from each trial. This is not straightforward in practice: reviewers must decide how to handle missing outcome data in the contributing trials.

OBJECTIVE

To investigate methods of allowing for uncertainty due to missing data in a meta-analysis.

STUDY DESIGN AND SETTING

The Cochrane Library was surveyed to assess current use of imputation methods. We developed a methodology for incorporating uncertainty, with weights assigned to trials based on uncertainty interval widths. The uncertainty interval for a trial incorporates both sampling error and the potential impact of missing data. We evaluated the performance of this method using simulated data.

RESULTS

The survey showed that complete-case analysis is commonly considered alongside best-worst case analysis. Best-worst case analysis gives an interval for the treatment effect that includes all of the uncertainty due to missing data. Unless there are few missing data, this interval is very wide. Simulations show that the uncertainty method consistently has better power and narrower interval widths than best-worst case analysis.

CONCLUSION

The uncertainty method performs consistently better than best-worst case imputation and should be considered along with complete-case analysis whenever missing data are a concern.

Is fetal gender significant in the perinatal outcome of pregnancies complicated by placental abruption?

Seventy cases of placental abruption were studied. These occurred in 12,800 deliveries, an incidence of 4.8 per 1000. There were more female infants (34-56%) compared with males (27-44%). Sixteen infants were stillborn and one baby died in the neonatal period. Thirteen of the perinatal deaths occurred in male infants. We speculate that a higher metabolic rate in male infants may account for this sex difference.

Pro-active call center treatment support (PACCTS) to improve glucose control in type 2 diabetes: a randomized controlled trial

OBJECTIVE

To determine whether Pro-Active Call Center Treatment Support (PACCTS), using trained nonmedical telephonists supported by specially designed software and a diabetes nurse, can effectively improve glycemic control in type 2 diabetes.

RESEARCH DESIGN AND METHODS

A randomized controlled implementation trial of 1-year duration was conducted in Salford, U.K. The trial comprised 591 randomly selected individuals with type 2 diabetes. By random allocation, 197 individuals were assigned to the usual care (control) group and 394 to the PACCTS (intervention) group. Lifestyle advice and drug treatment in both groups followed local guidelines. PACCTS patients were telephoned according to a protocol with the frequency of calls proportional to the last HbA(1c) level. The primary outcome was absolute reduction in HbA(1c), and the secondary outcome was the proportion of patients reducing HbA(1c) by at least 1%.

RESULTS

A total of 332 patients (84%) in the PACCTS group and 176 patients (89%) in the control group completed the study. Final HbA(1c) values were available in 374 patients (95%) in the PACCTS group and 180 patients (92%) in the usual care group. Compared with usual care, HbA(1c) improved by 0.31% (95% CI 0.11-0.52, P = 0.003) overall in the PACCTS patients. For patients with baseline HbA(1c) >7%, the improvement increased to 0.49% (0.21-0.77, P < 0.001), whereas in patients with baseline HbA(1c) <7% there was no change. The difference in the proportions of patients achieving a >/=1% reduction in HbA(1c) significantly favored the PACCTS intervention: 10% (4-16, P < 0.001) overall and 15% (7-24, P < 0.001) for patients with baseline HbA(1c) >7%.

CONCLUSIONS

In an urban Caucasian trial population with blood glucose HbA(1c) >7%, PACCTS facilitated significant improvement in glycemic control. Further research should extend the validity of findings to rural communities and other ethnic groups, as well as to smoking and lipid and blood pressure control.

Interventions for photodamaged skin

BACKGROUND

Photodamage describes skin changes such as fine and coarse wrinkles, roughness, freckles and pigmentation changes that occur as a result of prolonged exposure to the sun. Many treatments are available to reverse the damage, but it is unclear which work and at what cost in terms of unwanted side effects.

OBJECTIVES

To assess the effects of topically applied treatments, tablet treatments, laser and surgical procedures for photodamaged skin.

SEARCH STRATEGY

We searched the Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library, Issue 1 2002, MEDLINE (1966-June 2002), EMBASE (1974-June 2002), Health Periodicals (1976-June 2002). We checked references of articles and communicated with authors and the pharmaceutical industry.

SELECTION CRITERIA

Randomised controlled trials which compared drug or surgical interventions with no treatment, placebo or another drug, in adults with mild, moderate or severe photodamage of the face or forearms.

DATA COLLECTION AND ANALYSIS

Two reviewers independently extracted data and assessed trial quality.

MAIN RESULTS

Thirty studies of variable quality were included. Eight trials showed that topical tretinoin cream, in concentrations of 0.02% or higher, was superior to placebo for participants with mild to severe photodamage on the face and forearms (although losses to follow-up were relatively high in most studies). For example, the relative risk of improvement for 0.05% tretinoin cream, compared to placebo (three studies), at 24 weeks, was 1.73 (95% confidence interval 1.39 to 2.14). This effect was not seen for 0.001% topical tretinoin (one study) or 0.01% (three studies). A dose-response relationship was evident for both effectiveness and skin irritation. One small within-patient study showed benefit from topical ascorbic acid compared with placebo. Tazarotene (0.01% to 0.1%) and isotretinoin (0.1%) both showed significant improvement over placebo for moderate photodamage (one study each). There is limited evidence (one trial), to show that the effectiveness of 0.05% tretinoin, is equivalent to the effects of 0.05% and 0.1% tazarotene. One small study showed greater improvement in upper lip wrinkles with CO2 laser technique compared to Baker's phenol chemical peel, at 6 months. Three small RCTs comparing CO2 laser with dermabrasion found no difference in wrinkle score at 4 to 6 months, suggesting that both methods are equally efficacious, but more erythema was reported with the laser. The effectiveness of other interventions such as hydroxy acids and natural polysaccharides was not clear.

AUTHORS' CONCLUSIONS

There is conclusive evidence that topical tretinoin improves the appearance of mild to moderate photodamage on the face and forearms, in the short term. However erythema, scaling/dryness, burning/stinging and irritation may be experienced initially. There is limited evidence that tazarotene and isotretinoin benefit patients with moderate photodamage on the face: both are associated with skin irritation and erythema. The effectiveness of other interventions remains uncertain.