Pirie 2013.
| Methods |
Study design: RCT Location: Northern Ireland Setting: Royal Jubilee Maternity Service, Belfast, Northern Ireland Recruitment period: Februrary 2005 and December 2007 |
|
| Participants |
Inclusion criteria: females >18 years old with singleton pregnancy and ≥ 20 natural teeth Exclusion criteria: multiple pregnancy, diabetes/pregnancy complications, requiring antibiotic prophylaxis before periodontal scaling, had been provided with specialist periodontal treatment in the previous 12 months or aggressive periodontitis requiring urgent intervention Mean age (± standard deviation (years)): Group A = 30.5 ± 4.5, Group B = 30.5 ± 5.5 Mean gestational age (± standard deviation (days)): Group A = 97.6 ± 10.2, Group B = 98.8 ± 10.8 Previous preterm: Group A = 0, Group B = 1 Periodontitis: it was defined as ≥ 4 mm probing depth at ≥ 4 sites and clinical attachment level ≥ 2 mm at ≥ 4 sites Number randomised: n = 99 Number evaluated: n = 99 |
|
| Interventions |
A) Antenatal periodontal treatment ‐ SRP (n = 49): oral hygiene instruction, followed by supragingival and subgingival scaling and root planing of sites with probing depths ≥ 4 mm and polishing of all the teeth. Therapy was performed over 2 1‐hour sessions under local anaesthetic (9 patients refused anaesthetic due to anxiety). Treatment was completed by 24 weeks gestational age. No information on the expertise of the dental professional who administered intervention B) Antenatal periodontal treatment ‐ Alternative mechanical treatment (n = 50): oral hygiene instruction and supragingival cleaning of all teeth at their baseline visit and the option of postpartum periodontal treatment All women: periodontal examination by calibrated examiner. Post‐treatment clinical periodontal related data were collected at 8 weeks after treatment (32 weeks gestational age) |
|
| Outcomes | Pregnancy complications such as pre‐eclampsia, type of delivery, birth weight, gestational age, probing depth, clinical attachment loss, plaque, bleeding on probing | |
| Funding | Supported by Research and Development Office, Department of Health, Northern Ireland Grant EAT/2560/03 | |
| Notes | For sample size, 50 participants in each group were to achieve a power of 80% to detect a difference of 0.6 in mean birth weight standard deviation score equating to a difference of 300 g in birth weight | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Quote: "Allocations were computer generated by a third person who was not otherwise involved in the study" |
| Allocation concealment (selection bias) | Low risk | Quote: "Randomly allocated to either the control or test arm of the study using sealed opaque envelopes labelled with a study number and containing the group allocation" |
| Blinding of participants and personnel (performance bias) All outcomes | High risk | Blinding not feasible |
| Blinding of obstetric outcome assessment (detection bias) | Low risk | Birth outcomes were completed at delivery by delivery‐suite staff. These staff members were masked to the group assignments of the participants |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | Intention‐to‐treat principle applied |
| Selective reporting (reporting bias) | Low risk | All expected outcomes were reported |
| Other bias | Low risk | Both groups were balanced for age, weight, height, BMI, alcohol consumption, smoking, periodontal condition, obstetric history except for social class with high socioeconomic status in the test group compared to the control (P = 0.02). However, the review authors did not consider this to be sufficient to bias the results |