Table 1 Details of the seven randomised or controlled clinical trials where there has been stepwise, partial or no caries removal in primary teeth compared with conventional restorations (search strategy available on request)

Magnusson ³⁹ (1977)

Randomised parallel group study set in one secondary care site with four operators (Sweden)

62 children (510 years) 110 primary molars. Seem to be occlusal lesions only

Intervention

Stepwise (partial caries removal – with re-entry after 46 weeks); temporary – calcium hydroxide, intermediate layer of ‘Dropsin’ and zinc oxide eugenol cement.

Control

Complete caries removal (restorative material not stated)

100% follow up at 1 year

Pulp exposure during treatment

Intervention:

First stage: 0/55 (0%)

Second stage: 8/55 (14.5%)

Control:

29/55 (52.7%)

‘Consequently, judged by the clinical criteria used [stepwise caries removal] with a calcium hydroxide inlay may obviate a considerable number of pulp treatments in primary molars.’

Ribeiro ⁴⁰ (1991)

Randomised parallel group study (Brazil)

38 children (711 years) 48 primary molars. Equal Class I and Class II restorations carried out. Caries into dentine ‘at least 2 mm wide’

Intervention

Partial caries removal: removal of carious dentine from enamel-dentine junction (EDJ) but visible, moist, soft dentine not removed from floor or axial walls & immediate placement of definitive composite restoration.

Control

Complete caries removal and placement of composite restoration

100% follow up at 1 year

Signs/symptoms pulpal pathology

Intervention:

0/24 (0%)

Control:

1/24 (4.2%)

Longevity of restoration

Intervention & control:100% both arms

‘Application of an adhesive restorative system to irreversibly infected dentin did not affect the clinical performance of the restoration.’

Innes^13,26 (2007 and 2011)Pragmatic, multi-centre split mouth, RCT set in primary care with 17 operators – general dentists (Scotland)

132 children (310 years) 264 primary molars. Class I (33%) and Class II lesions/restorations (67%)

Intervention

Hall Technique with 42% of teeth caries radiographically >half way through dentine.

Control

Dentists usual treatment including caries removal 69% GI; 11% composite; 8% amalgam; 5% compomer; 1% PMC; 2% fissure sealant

94% (124/132) at 2 years and 69% (91/132) at 5 years

Signs/symptoms pulpal pathology

At 5 years: p = 0.000488; NNT 8 in favour of intervention.

Intervention:

2 yrs: 3/128 (2%)

5 yrs: 3/91 (3%)

Control:

2 yrs: 19/128 (15%)

5 yrs: 15/91 (16.5%)

Longevity of restoration

At 5 years: p <0.000001; NNT 3 in favour of intervention.

Intervention:

2 yrs: 6/128 (5%); 5 yrs:4/91 (5%)

Control:

2 yrs: 57/128 (46%); 5 yrs: 38/91 (42%)

‘…sealing-in caries by the Hall Technique statistically, and clinically, significantly outperformed the GDPs' standard restorations. Hall technique outcomes were comparable with those of standard restorations in studies in secondary care. These results strongly support the Hall technique as a predictable restorative option, with low failure and, therefore, re-treatment, rates for managing carious primary molars in a primary care environment.’

Lula ⁴¹ (2009)

Parallel group randomised control trial. Secondary care with multiple operators (Brazil)

30 children (58 years); convenience sample. 36 primary molars. Caries extending into inner half of dentine on radiograph; occlusal and occluso-proximal. Sometimes more than one tooth per child included

Intervention

Partial caries removal; microbiological samples taken; calcium hydroxide base; restored with composite.

Control

Complete caries removal; microbiological samples taken; calcium hydroxide base; restored with composite

1 year follow up. Children; 87% (26/30) Teeth; 89% (32/36) Four children and four teeth were lost to follow up, two from each arm

Bacterial growth from dentine samples

‘No difference in microbial growth between groups was observed after 36 months for any of the microorganisms studied.’

Pulp exposure during treatment

Intervention & control:

0% both arms

Longevity of restoration

Intervention:

16/16 (100%)

Control:

15/16 (94%)

‘The results suggest that persistence of bacteria does not seem to be a reason for reopening of cavities in deciduous teeth after partial caries removal.’

Orhan ⁴² (2010)

Parallel group randomised control trial (Turkey)

123 children (415 years) 94 mandibular second primary molars with caries extending >three-quarters through dentine radiographically. (Also included 60 mandibular permanent first molars)

Intervention

Group 1: Partial caries removal and compomer restoration.

Group 2: Stepwise caries removal if pulp exposure suspected – calcium hydroxide base, ZOE; re-entry after 3 months; restoration with glass-ionomer base and compomer.

Control

Complete caries removal and compomer restoration

1 year follow up Teeth; 78% (73/94)

Pulp exposure no statistically significant difference between partial and stepwise caries removal or between stepwise and complete caries removal.

Intervention:

Group 1: 2/31 (6.5%)

Group 2: 3/32 (9.4%)

Control:

6/31 (19%) Signs/ symptoms pulpal pathology (NB unexposed teeth only)

Intervention:

Group 1: 0/29 (0%)

Group 2: 1/29 (3%) (lost temporary then abscess 1/29)

Control:

2/25(8%) (Internal resorption 2/25)

‘Indirect pulp therapy in both primary and young permanent teeth can be used successfully with a 1 or 2 visit approach.’

Borges ⁴³ (2012)

Single centre randomised trial set in University Dental Centre (Brazil)

30 children (59 years)

Two unrestored, non-cavitated teeth with occlusal caries into dentine per child

Two arm RCT; each child had two teeth entered to the trial but not clear if one assigned to each arm

Intervention

Rubber dam isolation, cleaned and fissure sealant placed.

Control

Local anaesthesia, rubber dam isolation, high speed access to caries, 'carious tissue' removed and tooth restored with composite

1 year follow up. Children 87% (26/30) Clinical cariesprogression or cavitation in sealant group and radiographic progression

Radiographic lesion progression not statistically significant p = 0.12

Intervention:

3/26

Control:

0/26

Longevity of restoration not statistically significant p = 0.12

Intervention:

Complete retention 23/26 (88.5%)

Partial retention 3/26 (11.5%)

Complete sealant loss 0/26

Control:

Complete retention 26/26 (100%)

‘Fissure sealing and tooth restoration were equally effective in the management of non-cavitated dentine occlusal caries in primary teeth. Invasive procedures can be replaced with the non-drilling approach with no adverse consequences for paediatric patients.’

Phonghanyud ⁴⁴ (2012)

‘Two standard dental clinics in two hospitals’. Single operator. (Thailand)

276 children (611 years) Occlusal and/or proximal surface caries extending >one-third through dentine without signs/symptoms of irreversible pulpitis

Three arm RCT

Intervention

Group 1: Partial caries removal at EDJ – spoon excavation

Group 2: Complete caries removal – spoon excavation.

Control

Group 3: Complete caries removal – rotary instruments (LA used for five children). All cavities accessed with high speed round bur & teeth restored with glass-ionomer cement (GIC)

1 year follow up. Children 96% (266/276) clinical and radiographic

Cumulative survival rates of restorations not statistically significant for any groups

Group 1 83%,

Group 2 83%

Group 3 (Control) 89%

Pulp survival not statistically significant for any group

Group 1 99%,

Group 2 100%,

Group 3 (Control) 98%

However teeth excluded prior to analysis:

Group 2 – 1 pulp exposure

Group 3 – 2 pulp exposures

‘The clinical and radiographic evaluations after 12 months indicated that partial soft caries removal at EDJ followed by GIC restoration was comparable to that of ART and conventional approaches.’