Skip to main content

Table 3 Medications that increase emergency hospital admissions

From: Medications that reduce emergency hospital admissions: an overview of systematic reviews and prioritisation of treatments

Author, YearID No. Patient population Medication treatment Control Patients (RCTs) Mean agea Outcome mean follow-upb NNH (95% CI)
RR (95% CI)
I2% Quality of the evidencec
Increase hospital admissions from out-patient, day-procedure, or community settings
 Baigent, 20131653a Patients indicated for NSAID treatment COX2-inhibitors Placebo 88,367 (184) Unclear HF hospitalisation, Follow-up unclear NNH: 3 (2 to 6)
rr: 2.28 (1.62 to 3.20)d
Unclear Moderate
 Sampson, 20136806f Schizophrenia Intermittent antipsychotic therapy Maintenance antipsychotic therapy 661 (6) 35 ± 5 Hospitalisation at 21 months NNH: 7 (4 to 15)
RR: 1.58 (1.28 to 1.97)
19 High
 Kew, 20144004z Moderate to severe stable COPD Fluticasone Placebo 16,338 (15) 64 ± 1 Pneumonia hospitalisation at 12 months NNH: 164 (114 to 259)
RR: 1.81 (1.51 to 2.17)
0 High
  1. ID No. identification number, RCT randomised controlled trial, NNH number needed to treat to cause one emergency hospital admission, RR risk ratio, CI confidence interval, NSAID non-steroidal anti-inflammatory drugs, COX2 Cyclooxygenase 2, HF heart failure, rr rate ratio, COPD chronic obstructive pulmonary disease
  2. aMeans and standard deviations for patient ages in years. Mean of means and median ages in trials contributing to the pooled estimate. Not all included trials reported age in a form that could be averaged
  3. bMean length of follow-up across trials, unless otherwise specified. Mean of means, medians and total study durations reported in trials contributing to the pooled estimate. Not all included trials reported follow-up in a form that could be averaged. Rounded to the nearest whole month
  4. cModerate- or high-quality evidence according to the Grading of Recommendations Assessment, Development and Evaluation working group criteria
  5. dRate ratio