Patient smoking status and its impact on periodontal treatment efficacy

Background

Periodontitis is one of the most common diseases of the oral cavity of Norwegian habitants, a disease that may cause tooth loss if it remains untreated. Good dental hygiene is important for prevention of periodontitis, whereas established disease can be treated by dentist by using methods such as scaling and root planning (SRP), surgical flap procedures, gingivectomy, antibiotics or a combination of the foregoing.

Cigarette smoking is known as a factor that may negatively affect the oral health. As the circulation of the oral tissues usually is reduced in smokers, it has also been suggested that smoking is risk factor for development of periodontal diseases. Finally, it is hypothesized that the smoking habits of a patient may directly affect the outcome of professional periodontal treatment.

The Norwegian Knowledge Centre for Health Services was requested by the Norwegian Directorate of Health to perform a systematic literature review with respect to the relationship between the patient smoking habits and the effect of periodontal treatment offered by a dentist.

Method

A search strategy was developed to identify relevant RCT’s and cohort studies. The search was performed in Medline, Embase, Svemed and Cochrane Library (CENTRAL and CRD) on 28. May 2008.

The relevance of identified literature was established by evaluation against predefined inclusion and exclusion criteria. For example, studies were excluded if relevant effect measures (e.g. tooth loss, pocket depth or clinical attachment level) were not reported for smokers as well as non-smokers. The results of relevant research articles were summarized in tables, described in the text, and also in meta-analysis were this was appropriate.

Results

The search after literature return 561 unique references, but 495 were considered irrelevant after screening of abstracts. Full text versions of the remaining references were ordered and a total of 38 were considered to fulfil all inclusion criteria.

Tooth loss among patients undergoing periodontal therapy was reported in one of the included studies, and showed a statistically insignificant trend towards increased risk of tooth loss in the smoker group (Odds Ratio 2.27; 95 % Confidence Interval 0.86 to 5.94). The precision of the effect estimate was poor, probably reflecting a skewed distribution of smokers and non-smokers (26 versus 74), and a relatively few observed events of tooth loss. 

Changes in pocket depths and clinical attachment levels (measured by a periodontal probe) were reported in respectively 35 and 32 of the included primary studies. Outcome measurements were extracted and incorporated into meta-analysis, revealing large heterogeneities in and between the reported results. Various sources of heterogeneity were identified by the analysis, including pre-therapeutic pocket depth, different treatment procedures and differences in smoker definition criteria. Additionally, the precision of the effect estimate is impaired by the large methodical uncertainties associated with the use of periodontal probes, contributing low precision and broad confidence interval in individual studies.

The heterogeneity in the reported results complicates data interpretation, but in total it seems like non-smoking patients show better effect of periodontal treatment than smoking patients. Post-therapeutic improvement in pocket depth and clinical attachment level seems to be higher in non-smokers compared to non-smokers – both when pocket depth (Mean Difference 0.33; 95 % Confidence Interval 0.22 to 0.43 mm) and attachment levels (Mean Difference 0.30; 95 % Confidence Interval 0.19 to 0.41 mm) are used as outcome parameters.

The greatest difference in treatment efficacy between non-smokers and smokers was found in studies where pre-therapeutic pocket depths were greater than 7 mm (Pocket depth: (MD 0.87; 95 % CI 0.49 to 1.24 mm); attachment level: (MD 0.75; 95 % CI 0.33 to 1.18 mm)).

 Treatment efficacy among never-smokers was also compared to treatment efficacy among ex-smokers, and the results suggest that these two groups obtain comparable treatment efficacy, both with respect on pocket depth (MD 0.07; 95 % CI -0.09 to 0.23 mm) and clinical attachment level (MD 0.11; 95 % CI -0.10 to 0.32 mm).

Six studies reported changes in radiological attachment as separate outcome. All tended to show improved treatment effect in the non-smoker groups, but the differences were not statistically significant.

Discussion

In total, 34 of 38 included primary studies tended to show increased effect of periodontal treatment in non-smokers compared to smokers. The reported results were characterized by heterogeneity, implying that the treatment outcome depends on several factors. Such factors may include details associated with treatment procedures, cigarette consume (dose), pocket depth prior to therapy and others.

It is important to be aware that studies with randomised design are not available for this kind of research questions and thus, our knowledge is based on studies with observational designs. Based upon observational studies, it is a risk that the final result will be influenced by confounding factors. Thus, the relationship between smoking and treatment efficacy that is seen in this systematic review is not necessarily related to the action of smoking alone, but may reflect general differences in lifestyle between smokers and non-smokers (different diet, differences in personal oral hygiene etc.).

Nevertheless, taking into account that exposure to cigarette smoke has several well-known effects on the physiology of the oral cavity; it is not difficult to find scientific arguments that support a direct relationship between cigarette exposure and therapeutic effect.  

Conclusion

1. Tooth loss among smoking and non-smoking patients who received periodontal treatment was reported in one study. Results were characterized by poor precision, and no statistically significant difference was observed.
2. When reduced pocket depths and reduced attachment loss serve as outcome measures, post-therapeutic outlooks seem to be better for non-smokers than for smokers. However, it is not possible to determine whether these differences reflect actual differences in therapeutic response or differences in natural disease progression.
3. The effect of periodontal therapy among never-smokers and among patients who had quit smoking prior to therapy seemed to be comparable.

Patient smoking status and its impact on periodontal treatment efficacy is examined in several studies, but new research is still needed. First, large studies with many participants and standardized treatment are rare. Second, studies performing thorough dose-response analysis (i.e. correlating outcome measurements and actual cigarette consume for each patient).  

A weakness with the existing data material is the uncertainty that is related to assessment of pocket depth and clinical attachment loss. The periodontal probe has an uncertainty of approximately 1 mm, a number that appears to be large compared to the actual differences in treatment effect. Identification of more precise outcome measures would probably provide more accurate knowledge about the relationship between patient smoking habits and periodontal treatment efficacy.