Chapter  48:  Cardiovascular Effects of Epinephrine on Hypertensive Dental Patients: Evidence Report/Technology Assessment Number 48

Objectives

Hypertension affects one quarter of the adult U.S. population. A systematic review of the literature was conducted to identify the additional risks of adverse cardiovascular outcomes to controlled and uncontrolled hypertensive individuals represented by the use of epinephrine-containing anesthetic solutions and epinephrine-impregnated retraction cords during dental treatment.

Methods

Two parallel searches of MEDLINE, EMBASE, and the Cochrane Controlled Trials Register were conducted to identify reports involving cardiovascular and hemodynamic outcomes associated with the use of epinephrine in the dental treatment of hypertensive individuals. One search identified reports involving epinephrine in local anesthetic solutions, the other epinephrine in gingival retraction materials. The gray literature was not searched, and only reports in English were included in the search. We included any report where epinephrine was administered, and one or more cardiovascular parameters (blood pressure, heart rate, cardiac output, plasma epinephrine concentration, EKG changes including transient arrhythmias) or adverse events (headache, syncope, angina, hypertensive crisis, longer-term arrhythmia, cerebral vascular accident, myocardial infarction) was reported for hypertensive individuals. We selected studies for inclusion in the evidence table from among 373 local anesthetic and 33 retraction cord reports identified in the searches using independent dual review of titles, abstracts, and full papers. We abstracted data into the table for six reports for local anesthesia. No studies of retraction cord met the inclusion criteria.

Results

Five of the studies reported blood pressure and heart rate changes associated with use of epinephrine in local anesthetics in uncontrolled hypertensive patients. One study reported results among controlled hypertensives. We did not meta-analyze these results due to incomplete information describing variability. In general, hypertensive subjects receiving an extraction experience small increases in systolic blood pressure and heart rate associated with the use of a local anesthetic containing epinephrine (4 mm Hg and 6 bpm). These increases associated with the use of epinephrine occur in addition to increases in systolic and diastolic blood pressure and heart rate associated with undergoing the procedure without epinephrine (11.7 and 3.3 mm Hg and 4.7 bpm) that are larger for hypertensives than for normotensives. No adverse outcomes were reported in these studies, and only one case report of an adverse outcome associated with the use of epinephrine in local anesthetic in a hypertensive patient was identified in the literature. However, we judged the strength of the evidence for answering the question addressed by the review as poor. The available studies did not address effects of gingival retraction cord, and only one study examined effects of epinephrine in anesthetics on controlled hypertensive subjects, where additional risks due to interactions with antihypertensive medications are possible. Further, the five studies examining effects in uncontrolled hypertensive subjects examined a narrow range of outcomes, and their designs made it unlikely that three of the studies could detect transient changes in outcomes that might indicate increased risk for adverse events.

Conclusions

The quantity and quality of the pertinent literature is problematic. The reported occurrence of adverse events in hypertensives associated with the use of epinephrine in local anesthetics is minimal, and the increased risk for adverse events among uncontrolled hypertensives is low, according to the authors of the studies that have examined this issue. However, these studies, as a group, do not offer strong evidence for this conclusion because the ability of the studies to detect signs of increased risk, mainly transient changes in cardiovascular and hemodynamic parameters, was limited. Further, no information is available to quantify the risk of using epinephrine-impregnated gingival retraction cord in hypertensive individuals, or epinephrine-containing local anesthetics in controlled hypertensives.

Overview

The question addressed by this systematic review is whether there is evidence of added risk of adverse events posed by the hemodynamic and cardiovascular effects of epinephrine when it is used in the dental treatment of hypertensive individuals. About 24 percent of the U.S. adult population is hypertensive, when hypertension is defined by a mean systolic blood pressure of 140 mm Hg or greater, a diastolic blood pressure of 90 mm Hg or greater, or use of prescription antihypertensive medication. Of this hypertensive population, 48 percent are untreated, 24 percent are successfully treated, and 28 percent are inadequately treated. Thus, hypertensive individuals, both controlled and uncontrolled, will represent a substantial proportion of a typical dental practice's adult patients.

Epinephrine is widely used as an additive in local anesthetics (typically in concentrations of 1:100,000) to improve the depth and duration of the anesthesia, as well as to reduce bleeding in the operative field. Epinephrine counteracts the anesthetic's localized vasodilator effects in subcutaneous and submucosal vessels, thereby reducing the risk of anesthetic toxicity by decreasing the rate of systemic absorption from the site of injection. Epinephrine is also impregnated in cotton cord that is inserted into the sulcus between a tooth and the surrounding gingiva, improving access for tooth preparation and allowing dental impression material to more readily flow into the sulcus to record details of teeth prepared for crowns. The epinephrine also constricts the blood supply to adjacent tissue, thereby permitting the impression to be secured without contamination by bleeding.

Despite these benefits, which may not be as readily achievable through use of non-epinephrine preparations, the clinical significance of cardiovascular and hemodynamic changes caused by the introduction of exogenous epinephrine makes its use among hypertensive individuals a controversial subject in dentistry. The added risks attributed to the use of epinephrine in hypertensive patients include greater probability of acute hypertensive crisis, angina pectoris, and myocardial infarction, as well as cardiac arrthymias through the direct action of epinephrine; and acute hypertensive or hypotensive crisis brought about by the interaction of epinephrine and some antihypertensive medications.

Recommendations for the use of epinephrine in clinical dental practice are not in full agreement. Most recommendations advise caution in using local anesthetics with epinephrine in hypertensive patients. However, some indicate that epinephrine is contraindicated in controlled hypertensive patients taking medications with known epinephrine interactions, while others indicate that epinephrine use is acceptable with appropriate precautions and monitoring. Opinion about the use of epinephrine in uncontrolled hypertensives is also divided, with some cautioning against it, while others indicate that it is appropriate in most instances. Virtually all recommendations discourage the use of epinephrine-impregnated gingival retraction cord in uncontrolled hypertensives, including those of manufacturers.

Methods

The key question for this systematic review is stated as, "What additional risks of adverse cardiovascular outcomes do epinephrine-containing local anesthetic solutions and epinephrine-impregnated gingival retraction cords represent for controlled and uncontrolled hypertensive individuals receiving dental treatment?" Because adverse events are relatively rare, we examined the literature for studies reporting changes in risk indicators for adverse events, as well as for adverse events themselves. We defined adverse events as headache, syncope, angina, hypertensive crisis, longer-term arrhythmia, cerebrovascular accident, and myocardial infarction. We considered risk indicators to include changes in blood pressure, heart rate, and stroke volume, and plasma epinephrine concentration, and EKG disturbances including transient arrhythmias.

We conducted separate literature searches focusing on effects of epinephrine in anesthetic solutions and in gingival retraction cord. In both instances we searched MEDLINE initially, with additional searching conducted in EMBASE and the Cochrane Controlled Trials Register. No attempt was made to search the gray literature, i.e., dissertations, theses, unpublished studies, abstracts, industry reports, and other nontraditional sources. We limited the searches to English-language reports. Subsequently, we examined reference lists of studies identified in these searches to include additional reports of possible interest. We identified 373 reports addressing the use of epinephrine-containing local anesthetics and 33 addressing epinephrine-impregnated gingival retraction cord. We then reviewed these studies for possible inclusion in the evidence table. We used essentially identical inclusion criteria in both reviews that addressed the inclusion and separate analysis of known hypertensive subjects, exposure to known concentrations of epinephrine through receipt of an intraoral injection or application of gingival retraction cord, recording of at least one cardiovascular or hemodynamic outcome, and a dental setting with dental treatment provided. We used independent dual review and eventually identified six local anesthetic studies that met the criteria for inclusion. No retraction cord papers met the criteria because no studies included hypertensive subjects.

We abstracted data from the included studies directly into the evidence table. We did not meta-analyze the results because two of the studies reported no information describing variation about the mean. We rated the quality of each included study using a rating scale that assessed several elements of internal and external validity, including sample size, presence of a comparison group of normotensive subjects, use of control groups (without epinephrine), outcomes reported, measurement methods, statistical testing, determination of hypertensive status, and reporting issues. We then graded the strength of the combined evidence, using a three-category system. The evidence was considered to be good if the numbers of studies and subjects were large (10 or more studies, 500 or more subjects); the quality of the studies was generally high (median quality score of 70 or higher); the results of these studies were consistent; and, taken together, the results were comprehensive with respect to risks examined. The evidence was considered to be fair if the numbers of studies and subjects were adequate overall (5 or more studies, 200 or more subjects); the quality of the studies was generally acceptable (median of 55 or higher); the results of these studies were reasonably consistent, with inconsistencies reflected as quantitative rather than qualitative differences; and the principal known risks were adequately examined. The evidence was considered to be poor if the numbers of studies and/or subjects were small (fewer than 5 studies or 200 subjects) or the quality of the studies was generally low (median of less than 55); or there were substantial inconsistencies in the results, or the risks examined among the studies did not represent a reasonably complete assessment of known risks.

Results

The six included studies comprised 325 subjects, of whom 177 were identified as hyper-tensive. Of these, 14 percent were identified as taking medication for control of hypertension. In all studies the local anesthetic involved was 2 percent lidocaine, and epinephrine concentrations were divided between 1:100,000 (n=3 studies) and 1:80,000 (n=3 studies). Quantities of anesthetic solution injected were reported in four studies, with means ranging from 2 ml to 4.5 ml. The outcomes examined in these studies consisted principally of systolic and diastolic blood pressures and heart rate. EKG recordings were collected in two studies. The dental procedure involved was tooth extraction in five of the six studies, and "minor oral surgery" in the sixth.

The results suggest that hypertensive subjects receiving an extraction experience small increases in systolic blood pressure and heart rate associated with the use of a local anesthetic containing epinephrine (4 mm Hg and 6 bpm). These increases associated with the use of epinephrine occur in addition to increases in systolic and diastolic blood pressure and heart rate associated with undergoing the procedure without epinephrine (11.7 and 3.3 mm Hg and 4.7 bpm) that are larger for hypertensives than for normotensives. No adverse outcomes were reported among any of the subjects in the studies included in the review, and only one report of an adverse event associated with the use of epinephrine in local anesthetic in a hypertensive patient was identified in the literature.

We rated the strength of the evidence as poor for describing additional risks among controlled and uncontrolled hypertensives due to epinephrine-containing local anesthetic solutions and gingival retraction cords. These ratings result from both the number of available studies and their quality. For outcomes of the administration of local anesthetic solutions containing epinephrine to patients taking medications for the control of hypertension, one study comprising 14 subjects and two medications was available. Two other studies included patients taking antihypertensive medications, but outcomes were not reported separately. No studies described outcomes of the use of gingival retraction cord either for hypertensive patients or for those taking medications for the control of hypertension. There were five studies addressing outcomes of the use of epinephrine-containing anesthetic solutions in hypertensive patients. The strength of this evidence was rated as poor because the outcomes considered in the studies did not represent a reasonably complete assessment of risk indicators, and because transient effects in blood pressure and heart rate, the principal outcomes reported, might have remained undetected in three of five studies.

Future Research

Based on the available evidence, which suggests that adverse outcomes among hypertensive patients are infrequent and that hemodynamic outcomes, which may be viewed as risk indicators, reflect only minimal change, replication of existing studies does not represent an efficient method to further our knowledge of the risks for adverse cardiovascular outcomes associated with use of local anesthetics containing epinephrine. Rather, a large-scale descriptive study of adverse outcomes of the use of epinephrine-containing local anesthetics would seem to be indicated. A long-term protocol initiated in one or more large dental clinics that involves electronic capture of pre-existing cardiovascular diagnoses and medication status of all patients, together with information describing all adverse outcomes occurring during treatment, could begin to quantify the magnitude of additional risk represented by the use of epinephrine in hypertensive dental patients with minimal outlay of effort and expense. Only if the results of such an investigation indicate that the added risk is greater than deemed acceptable would additional trials to develop more sensitive methods for identifying patients at increased risk be justified.

With respect to the use of epinephrine-impregnated gingival retraction cord, studies are needed to quantify the absorption of epinephrine from gingival tissues. The effects of time, tissue condition, cord construction, and epinephrine concentration on plasma concentration of epinephrine should be determined in these studies. Once a better understanding of the possible range of epinephrine concentrations is gained, the risks associated with the use of these cords in hypertensive patients can be evaluated. At present, a single human study reports absorption levels.

Key Question and Causal Pathway

The key question for this review focuses on whether there are added risks for adverse effects posed by the hemodynamic effects of the epinephrine found in local anesthetics and in gingival retraction cord among hypertensive individuals receiving dental treatment. The key question is stated as: What additional risks of adverse cardiovascular outcomes do epinephrine-containing anesthetic solutions and epinephrine-impregnated retraction cords represent for controlled and uncontrolled hypertensive individuals receiving dental treatment?

The causal pathway (Figure 1) illustrates the context of the question in the dental practice environment. Epinephrine is used in anesthetic solutions to decrease local blood flow, thereby lengthening the duration and deepening the anesthetic effect while possibly lessening the risk of systemic toxicity due to circulating anesthetic solution. Epinephrine is used in gingival retraction cord to minimize gingival bleeding that can result from tooth preparation, and to cause the free gingiva to shrink away from the tooth surface, which facilitates securing an accurate impression of the prepared tooth. Epinephrine may pose an increased risk of adverse cardiovascular outcomes to individuals with hypertension both through the direct action of epinephrine on cardiovascular physiology (Q1_a), and through its action on the altered physiology created by various medications prescribed for hypertension control (Q1_b). Because hypertension is often a marker for other cardiac disease, risk also may be elevated due to decreased ability of the heart to compensate for epinephrine-induced change. Finally, the cardiovascular effects of epinephrine may be enhanced by aspects of the dental environment itself, which may increase plasma concentrations of catecholamines.

Literature Search

Two separate literature searches were conducted for the evidence report, one for local anesthetic solutions and one for gingival retraction cord. In both instances MEDLINE was searched initially, with additional searching conducted in EMBASE and the Cochrane Controlled Trials Register. No attempt was made to search the gray literature, i.e., dissertations, theses, unpublished studies, abstracts, industry reports, and other nontraditional sources. This decision was based on resource limitations and on previous experience at the RTI/UNC EPC indicating that such searches were generally unproductive.

Tables 1 and 2 show the MeSH terms used for searching MEDLINE, the principal databases for both searches, as well as the results of these searches. The searches were run in March 2001. The search terms do not reflect the entirety of the inclusion and exclusion criteria developed for the questions (see next section) as indexing tended to be unreliable and nonspecific through the 1980s, when much of the literature of interest was reported. Only at the broadest level could the search be limited, namely to human studies reported in English. For anesthetic solutions, all papers with keywords associated with local anesthesia and with either epinephrine or hypertension keywords were initially identified. For retraction cords an equally broad strategy was employed, identifying all reports with keywords for hypertension or epinephrine and for various permutations of the term "gingival retraction cord."

Tables 3 and 4 show the results of similar searches in EMBASE, using the same search strategy, with more general terms substituted where necessary. All studies not duplicated in the previous MEDLINE search are shown on the last line of these tables. Finally, searches of the Cochrane Controlled Trials Register identified no additional studies.

A supplemental search strategy involved perusal of the reference listings of studies identified in the searches. This strategy identified two types of additional studies, those published prior to 1966, the earliest date for inclusion of dental studies in MEDLINE, and studies with incorrect or incomplete index terms. This supplementation was fruitful in that two of the studies included in the evidence table were identified in this manner.

Inclusion and Exclusion Criteria

Table 5 shows the inclusion and exclusion criteria applied to the articles identified in the searches. Essentially identical criteria were used in both reviews, including the inclusion and separate analysis of known hypertensive subjects, exposure to known concentrations of epinephrine, receipt of epinephrine-containing local anesthetic via either intraoral submucosal block or infiltration injection or receipt of epinephrine-impregnated gingival retraction cord, recording of at least one cardiovascular or hemodynamic outcome, and a dental setting with dental treatment provided. Given the nature of the key question, the requirements for hypertensive subjects, epinephrine, and cardiovascular/hemodynamic outcomes are obvious. We accepted a variety of hemodynamic and cardiovascular outcomes due to the broad wording of the key question regarding risk indicators for adverse outcomes. We considered risk indicators to include changes in blood pressure, heart rate, and stroke volume, and plasma epinephrine concentration, and EKG disturbances including transient arrhythmias. We included reports of headache, syncope, angina, hypertensive crisis, longer-term arrhythmia, cerebrovascular accident, and myocardial infarction as adverse events. We imposed the requirement that studies be conducted in a dental environment while dental treatment was provided to ensure that the results of the studies, and hence the review, would have external validity. Because receipt of dental injections and subsequent dental treatment result in increased plasma concentrations of catecholamines in some individuals that may mimic or heighten the effects of exogenous epinephrine,⁷ it is important that this environment be duplicated in studies of epinephrine's effects if the key question is to be addressed directly. Finally, studies reporting on fewer than five hypertensive subjects were excluded to minimize bias resulting from case selection. Such case reports were identified and considered separately in the analysis.

Review of Identified Papers

We performed initial surveys of the titles of the articles included in both searches, identifying those articles for which the title indicated some possibility that the study was relevant and might satisfy the inclusion criteria. These surveys were performed independently by the Clinical and Research Directors of the dental component of the EPC, with all titles identified by either placed on the potentially eligible list. Forty-four percent of articles placed on the local anesthetic list were selected by both Directors, with 27 percent and 29 percent identified by the Clinical and Research Director, respectively. For gingival retraction cord, the percentages were 40 percent by both, and 40 percent and 20 percent by the Clinical and Research Director, respectively.

We then reviewed the abstract and, if needed, the full paper for all articles surviving the title review to develop a refined list of potentially eligible articles. From this refined list, a final review was independently performed by the two Directors to determine eligibility. The agreement between Directors was 100 percent for the final include/exclude decisions for all articles for both reviews. All included articles were identified in the initial survey of titles by both Directors.

Table 6 shows the numbers of articles remaining after each stage in the review process. The final numbers of papers included in the reviews were six and none, respectively, for anesthetic solutions and retraction cord. Reasons for exclusion for the 16 local anesthetic and 6 retraction cord papers excluded in the final review are shown in Tables 7 and 8.

Development of the Evidence Table

The evidence table was designed to report the most salient elements of the included studies while remaining manageable in terms of size and complexity. The table was completed by direct abstraction from the included articles without the use of abstraction forms. The research director performed the abstraction, and the Clinical Director reviewed the tables against the individual studies. All discrepancies were resolved by consensus.

The evidence table reports a quality rating for each included article. This rating was developed specifically for the review, but the approach reflects that used in all ratings developed by the RTI/UNC EPC.⁸ A set of quality items was specified that assessed several elements of internal and external validity, including sample size, presence of a comparison group of normotensive subjects, use of control groups (without epinephrine), outcomes reported, measurement methods, statistical testing, determination of hypertensive status, and reporting issues (Table 9). Item weighting was determined empirically, again with reference to assessment instruments previously developed by the EPC. A total of 20 points were available for 13 items, with partial scores possible for multipoint items. Raw scores were rescaled to a 0 to 100 scale. The included studies were assessed by a single rater using a standardized form. The resultant scores were not used as a part of the inclusion/exclusion decision. Rather they represented one of the considerations for grading the available evidence.

Grading the Evidence

The strength of the assembled evidence was graded in terms of the extent to which it answered the key question; that is, the extent to which it identified and quantified any additional risks for hypertensive patients from exposure to epinephrine in dental practice. Three rating categories were possible:

• Good: the numbers of studies and subjects are large (10 or more studies, 500 or more subjects); the quality of the studies is generally high (median 70 or higher); the results of these studies are consistent; and, when taken together, the results are comprehensive with respect to risks examined.

• Fair: the numbers of studies and subjects are adequate overall (5 or more studies, 200 or more subjects); the quality of the studies is generally acceptable (median 55 or higher); the results of these studies are reasonably consistent, with inconsistencies reflected as quantitative rather than qualitative differences; and the principal known risks are adequately examined.

• Poor: the numbers of studies and/or subjects are small (fewer than 5 studies or 200 subjects); or the quality of the studies is generally low (median below 55); or there are substantial inconsistencies in the results or the risks examined among the studies do not represent a reasonably complete assessment of known risks.

The grading criteria are subjective, and they reflect the general approach to grading the evidence used by all EPCs. The evidence was graded independently by the Research and Clinical Directors. Separate grades were assigned to the evidence addressing the use of epinephrine in local anesthesia and gingival retraction cord in uncontrolled and controlled hypertensives. If it had been necessary, agreement on a grade would be achieved by discussion leading to consensus. No discussion was necessary as the Directors' grades were identical for all applications.

Summary of the Studies

The first of the included studies³¹ compared eight groups of subjects, four normotensive and four hypertensive. Other cardiac conditions for these subjects were not reported. Within these groups, injections either did or did not contain epinephrine, and subjects either were or were not premedicated. Blood pressure and heart rate were recorded immediately before the injection, 5 minutes post-injection, and 10 minutes post-injection. The timing of the extraction was not indicated. Mean increases in systolic blood pressure were small and occurred principally in the first 5 minutes, with means of 3 to 7 mm Hg among hypertensive groups, and 1 to 5 mm Hg among normotensives. Pressures then declined from 5 to 10 minutes post-injection in all but the nonsedated normotensive group receiving epinephrine. None of the changes were statistically significant from baseline. No between-group changes were tested statistically. Smaller changes in diastolic pressures were observed with decreases in five groups of between 1 and 3 mm Hg, and increases in three groups of between 0 and 2 mm Hg at 5 minutes. Additional small changes were seen in the period 5 to 10 minutes post injection, with maximum changes over 10 minutes never more than 3 mm Hg. Again, changes were not significantly different from baseline, and no intergroup comparisons were reported. Heart rate changes showed much the same pattern as systolic pressure, with mean increases of 1 to 7 bpm among hypertensive groups, and 1 to 4 bpm among normotensive groups for baseline to 5 minutes, with still smaller, mixed changes in the 5 to 10 minute period. The authors concluded that neither epinephrine nor hypertensive status resulted in significantly different mean changes in blood pressure or heart rate following injection with local anesthetic (p>0.05). Medication status among hypertensive patients was not reported.

The second study compared 20 hypertensive patients, many of whom also had other cardiovascular disease, with a small group of six normotensive subjects with no cardiac complications.³² Both groups received extractions under local anesthesia with and without epinephrine at separate visits. Here, hypertensive subjects exhibited larger increases in systolic pressure, with means of 28 mm Hg and 21 mm Hg with and without epinephrine, compared to normotensive subjects, with increases of 13 mm Hg and 12 mm Hg with and without epinephrine, respectively. Roughly equal increases were associated in the epinephrine groups with injection and extraction, and these were additive. Diastolic changes were smaller and mixed in direction, and no clear hypertensive or epinephrine patterns emerged. Heart rate changes were most pronounced in hypertensive subjects receiving epinephrine. The authors indicate that changes from baseline and differences in standard deviations among groups were not statistically significant for all comparisons, but the specific comparisons made, tests used, and p values are not described. No alterations in the cardiac complex (S-T segment alteration, T-wave inversion, abnormal QRS complex) were identified. However, premature ventricular contractions (PVCs) were present and reported to be less frequent in the absence of epinephrine.

The third study consisted of two groups, hypertensive subjects and normotensive subjects, each of whom received extractions with three different local anesthetic solutions, one with epinephrine, one with norepinephrine, and one without either.³³ Other cardiac conditions were not reported. The norepinephrine results are not reported in the evidence table. Among all subjects, changes in blood pressure and heart rate responses were small, but hypertensive responses for systolic blood pressure and heart rate tended to be somewhat greater. Comparison of epinephrine versus no epinephrine anesthetic in both hypertensive and normotensive groups showed that diastolic decreases were greater among epinephrine groups in the period following the extraction procedure (p <0.05). Significant differences in systolic pressure and heart rate occurred in the hypertensive group, and these differences involved isolated readings following extraction. No side effects were attributed to epinephrine.

The fourth study included normotensives (n=15), hypertensives (n=9), and a third group, moderate hypertensives (n=15), who had systolic and diastolic blood pressures ranging between 140 and 160 mm Hg and 90 and 95 mm Hg, respectively.³⁴ Ten subjects had previous myocardial infarctions and seven had arrhythmias. One patient was reported to be taking a beta blocker and was excluded from analysis of heart rate. Blood pressure and heart rate were recorded at 5-minute intervals throughout visits for tooth extractions. An EKG was run for the immediate period of the injection and the extraction. Group means for maximum and minimum blood pressures recorded were reported. All patients but two received local anesthetic containing epinephrine. Heart rate changes were not analyzed by hypertensive status. Increases in mean maximum systolic pressure were similar among the three groups, ranging from 19 to 27 mm Hg. Similarly, increases in mean maximum diastolic pressure ranged from 11 to 20 mm Hg across groups. These increases were statistically significant compared to baseline values (p<0.01). Intergroup differences were not tested for significance. EKG disturbances were reported in two subjects, a sinus arrhythmia with multifocal supraventricular premature contractions in a hypertensive subject and a 10-second sinus arrest in a normotensive subject, presumed to be an "extreme vasovagal reaction."

The fifth study included 14 subjects diagnosed as hypertensive and taking control medications, either beta blockers or nonpotassium-sparing diuretics.³ Blood pressure and heart rate were recorded twice; once before the injection, and once 10 minutes postinjection. Plasma potassium concentrations were recorded at these same times. Differences between groups in change in blood pressure were not significant, with systolic increasing 1 and 3 mm Hg, while diastolic decreased 3 and 12 mm. Heart rate changes were significantly different (p<0.02), decreasing 3 bpm in the beta blocker group and increasing 7 bpm in the diuretic group. The plasma potassium concentration decreased significantly in the diuretic group compared to baseline (p<0.01) and to the beta blocker group (p<0.05).

The sixth study compared 18 patients with essential hypertension receiving an extraction and an equal number of matched controls, with the matching criteria not identified.³⁵ All patients received epinephrine. Of the 18 hypertensives, 10 were receiving antihypertensive treatment (n=2 diuretics; n=5 calcium channel antagonists; n=2 ACE inhibitors; n=1, ACE inhibitor and calcium channel blocker). Blood pressure and heart rate were recorded every two minutes and values averaged for treatment phases. From EKG monitoring, power spectral densities were computed, and the ratio of the low frequency band to the high frequency band was calculated (LF/HF ratio), along with the high frequency band as a proportion of the total frequency band (%HF). The patterns of change in blood pressure were similar for the two groups, with inter-group differences not tested for statistical significance. Changes in systolic pressures were all significantly different than baseline (p<0.05). Only normotensive subjects exhibited a significant increase in diastolic pressure from baseline, and only for the extraction phase of treatment. The LF/HF ratio declined significantly during anesthesia and extraction phases for hypertensives but not normotensives, suggesting suppression of the cardiac sympathetic nervous system. Percent HF declined significantly during anesthesia for normotensives, but not hypertensives.

Table 10 summarizes the mean maximum blood pressure and heart rate changes (from mean baseline values) reported in the three studies of uncontrolled hypertensives where both hypertensives and normotensives received injections with and without epinephrine.^31-33 Because two of the studies did not report confidence intervals for their observations, meta-analysis was not performed.^31,32 The unweighted means of the individual study means shown in Table 10 indicate that when dental extractions were performed using local anesthetic with epinephrine, systolic blood pressure was 4 mm Hg higher for hypertensive subjects and no higher for normotensive subjects than when the same procedures were performed using local anesthetic without epinephrine. Heart rate was higher with epinephrine than without for both hypertensives and normotensives (6 bpm and 5.6 bpm, respectively), while diastolic blood pressure was lower (1.0 and 4.7 mm Hg, respectively, for hypertensives and normotensives). The table also indicates that receiving the procedure without epinephrine is also associated with increases in blood pressure and heart rate compared to baseline and that, for systolic blood pressure and heart rates, these increases are greater in hypertensives than in normotensives (11.7 versus 5.0 mm Hg, and 4.7 bpm versus 0.7 bpm). Increases in diastolic pressure were similar at 3.3 and 4.0 mm Hg. The mean maximum changes in blood pressure tended to be greater for both hypertensives and normotensives in the two studies of uncontrolled hypertensives not included in the table.^34,35 Only one of these studies reported heart rate values, which were lower.³⁵

Where timing of the maximum reading was reported, it occurred in conjunction with the surgical procedure rather than the injection in almost all instances. However, for both hypertensives and normotensives receiving local anesthetic with epinephrine, smaller increases in systolic blood pressure and heart rate were associated with the injection, whereas these increases were not present for injections without epinephrine.

None of the studies noted any patient-reported adverse events. Examination of EKGs in two studies indicated occurrences of PVCs associated with epinephrine in the local anesthetic, and in one study these PVCs were also associated with hypertensive status.

In addition to examining the studies included in the review for descriptions of adverse outcomes, we also examined several reports of adverse events possibly involving either epinephrine or hypertensive patients among the persons identified in the literature search. Of eight reports of adverse events associated with local anesthesia that we examined, two were associated with vasopressors other than epinephrine,^36,37 and five concerned putative reactions to epinephrine in normotensive subjects.^38-42 One report described an episode of atrial fibrillation extending 10 minutes during tooth extraction following injection of 2 percent lidocaine with 1:200,000 epinephrine, 3.6 + 0.9 ml, in a subject with controlled hypertension (nifedipine) and cardiac disease.⁴³

Finally, several studies not included in the systematic review reported outcomes associated with one or more hypertensive patients receiving local anesthetics with epinephrine, but these patients were not analyzed as a group.^{10,12,13,16,17,21-23} In none of these studies were adverse outcomes reported that could be associated with groups including hypertensive patients.

In summary, the information from the available studies of uncontrolled hypertensive subjects suggests that hypertensive subjects receiving an extraction experience small increases in systolic blood pressure and heart rate associated with the use of a local anesthetic containing epinephrine (4 mm Hg and 6 bpm). These increases associated with the use of epinephrine occur in addition to increases in systolic and diastolic blood pressure and heart rate associated with undergoing the procedure without epinephrine (11.7 and 3.3 mm Hg and 4.7 bpm) that are larger for hypertensives than for normotensives. No adverse outcomes were reported among any of the subjects in the studies included in the review, and only one report of an adverse event associated with the use of epinephrine in local anesthetic in a hypertensive patient was identified in the literature.

Strength of the Evidence

For purposes of grading the strength of the evidence, the key question addressed by the systematic review was deconstructed into four separate analysis questions, each addressing a different dimension of the key question. These four questions addressed separately the risks associated with the use of epinephrine in anesthetic solutions and the risk associated with the use of epinephrine in gingival retraction cord; and these two uses of epinephrine were addressed separately for two different populations, hypertensives taking medication to control their blood pressure, and uncontrolled hypertensives with elevated blood pressure. We rated the strength of the evidence as 'poor'for each of these four questions. For three of the questions, use of gingival retraction cord containing epinephrine in controlled and in uncontrolled hypertensives, and the use of epinephrine in local anesthetic solutions in controlled hypertensives, this 'poor' rating results from the number of available studies. One study examined outcomes of the administration of local anesthetic solutions containing epinephrine to patients taking medications for the control of hypertension comprising 14 subjects and two medications. Two other studies included patients taking antihypertensive medications, but outcomes were not reported separately. There were no studies describing outcomes of the use of gingival retraction cord either for hypertensive patients, or for those taking medications for the control of hypertension.

There were five studies addressing the fourth question, outcomes of the use of epinephrine-containing anesthetic solutions in hypertensive patients. These studies included 311 subjects, 163 of whom were hypertensive. The mean quality score for the studies was 54, the median was 58. These aspects of the available evidence meet the criteria for a strength rating of 'fair.' However, the strength of the evidence must be rated as 'poor' because the outcomes considered in these five studies do not represent a reasonably complete assessment of risk indicators. In essence, only changes in blood pressure and heart rate received scrutiny in all five studies, and the comprehensiveness of these assessments was quite varied. Only one study specifically noted the presence or absence of patient-reported side effects. Two studies monitored single-lead EKGs for the detection of arrhythmias. One study used two-lead EKG for analysis of sympathetic and parasympathetic activity. No other outcomes were assessed. Among the blood pressure and heart rate monitoring methods, three studies collected initial epinephrine exposure readings 5 minutes postinjection, and the remaining two collected initial readings at 2 minutes and @@@frac12@@@ minute postinjection, respectively. Thus, transient effects in principal outcomes might have remained undetected in three of five studies.

Limitations of the Evidence Base

Three principal limitations of the database are clearly 1) the incomplete nature of the evidence describing the effects of intraoral local anesthetics among uncontrolled hypertensives 2) the lack of studies describing the effects of intraoral injection with local anesthetics containing epinephrine on hypertensive subjects taking medications for the control of hypertension, and 3) the lack of studies regarding the effects of epinephrine-impregnated gingival retraction cord in uncontrolled and controlled hypertensive subjects.

Available studies of the use of epinephrine-containing local anesthetics in dental treatment for hypertensive patients are limited due to the narrow range of outcomes examined and the incomplete nature of the assessments within the narrow range of outcomes. While changes in systolic and diastolic blood pressure and heart rate were evaluated, the studies generally failed to assess other important outcomes that may indicate increased cardiovascular risk such as stroke volume, plasma epinephrine concentration, and disturbances in cardiac activity indicative of arrhythmias or transient ischemia. Due to the complex interaction of factors that determine hemodynamic outcomes, potentially adverse cardiovascular system responses to epinephrine contained in local anesthesia solutions may not be detectable through discontinuous recording of blood pressures and heart rate. Discontinuous recording such as that used in most of the studies included in the review may also mask transient changes in these basic outcomes that mark heightened risk.

Our inability to include more than one study examining the effect of epinephrine-containing local anesthetic solutions among controlled hypertensives is also a serious limitation of the evidence base. Hypertensive medications are among the most commonly prescribed medication groups and, among the types of medications prescribed, at least two pose the potential for adverse outcomes. Interaction of nonselective beta blockers and epinephrine can lead to reduction in cardiac output through alpha receptor-induced elevation in blood pressure with concomitant compensatory vagal reflex-mediated reduction in heart rate.⁴⁵ As shown in the study included in the review, when epinephrine is used in patients taking nonpotassium-sparing diuretics, potassium levels can decrease.³ Because studies performed in the dental environment of patients taking these medications have not been reported, knowledge of outcomes and possible adverse events must necessarily be inferred from other studies in other environments.

Although we found no studies that examine the effects of epinephrine-impregnated gingival retraction cord in hypertensive subjects, several studies have reported the effect of retraction cord in normotensive subjects.^25,27-30 In general, mean effects on blood pressure and heart rate were minimal. However, only one of these studies measured plasma epinephrine concentrations, and this study did not evaluate the effect of lacerations in gingival tissue, which are assumed to be associated with increased uptake of epinephrine.³⁰ Better quantitative knowledge of plasma concentrations of epinephrine resulting from use of impregnated cord in healthy, lacerated, and inflamed gingival tissue, together with knowledge of associations between plasma concentrations and length of exposure, would be useful in improving our understanding of the possible risks to hypertensive subjects without necessitating extensive trials among this population.

An additional consideration with respect to limitations of the database may reflect more on the restrictions imposed by the key question. Part of the reason for caution in administering epinephrine to patients with hypertension lies in the fact that in middle-aged and older patients, the presence of hypertension increases the likelihood that other cardiovascular disease may also be present, which in turn reduces the resiliency of the system to challenges imposed by exogenous epinephrine. This review did not include studies of patients with cardiovascular disease unless a specific diagnosis of hypertension was included in the description for all subjects of at least one arm of the study. Several of the studies listed in Table 9 examine the effects of epinephrine on patients with other forms of cardiovascular disease. As a general observation, few patients in the included studies were identified as having the cardiac pathology described in the excluded studies.

Only four of the studies identified the range of anesthetic dosages that subjects received, and no dose-response analyses were reported. Dosages were all between 2 and 4.5 ml, or less than three full capsules. This information does not permit any extrapolation to higher doses, which might be attempted to extend time or depth of anesthesia. Thus, the only information available describing the effects of epinephrine administration on hypertensives covers only the most conservative dosing strategy.

Two final limitations of the evidence base are the relatively small number of hypertensive subjects included in the reviewed studies, and the tendency for studies to aggregate individual patient data, often without reporting any measure of variability, let alone assessment of outliers. As suggested by the results of these studies as well as by the paucity of reports of specific adverse outcomes, such adverse outcomes of dental anesthesia among hypertensive patients are uncommon events. Thus detecting changes in indicators that may signal heightened risk for adverse outcomes requires observations of large numbers of subjects and careful evaluation of variability among the observations.

Appendix A. Acknowledgements

This study was supported by Contract No. 290-97-0011 Task Number 6 from the Agency for Healthcare Research and Quality (AHRQ), formerly the Agency for Health Care Policy and Research (AHCPR). We acknowledge the assistance of Jacqueline Besteman, J.D., M.A., the AHRQ Task Order Officer for the Evidence-based Practice Center Program; and Ernestine (Tina) Murray, R.N., M.A.S., the AHRQ Task Order Officer for this task. The cooperation and support of Isabel Garcia, D.D.S., M.P.H., the National Institute of Dental and Craniofacial Research (NIDCR) liaison with AHRQ, has been greatly appreciated. In addition, the investigators appreciate the time and assistance of Ms. Terri Kissiah, Ms. Carrie Roseamelia, and Mr. Christian Evensen of RTI, and Dr. Lynn Whitener of UNC.

Appendix B. Selection and Role of the Technical Expert Advisory Group

We gratefully acknowledge the involvement of and assistance from the three members of our Technical Expert Advisory Group (TEAG). TEAG members are listed at the end of this Appendix. The TEAG was meant in part to contribute to (a) advancing AHRQ's broader goals of creating and maintaining "science partnerships" and "public-private partnerships" and (b) meeting the needs of a broad array of potential consumers and users of its products. Thus, it was both a substantive resource and a sounding board during the study. The TEAG is the body from whom "expert inputs" were formally sought at several points through the project.

We constituted our TEAG from two types of technical experts. These types are (1) technical/clinical experts and (2) potential users of the systematic review, including explicitly a representative of the primary dental professional organization in the U.S. -- the American Dental Association (A.D.A).

The final decision about TEAG membership was based on the candidates' availability for scheduled conference calls and other input, willingness to review materials and provide advice and assistance within a short turnaround time, and approval by the AHRQ TO Officer.

The RTI-UNC team solicited the views of TEAG members from the start of this systematic review. Among other issues, TEAG members provided insights and reactions to the key clinical question, input to the literature review process by ensuring that we included all known published research meeting our inclusion criteria, and reviewed a preliminary draft of this review.

In keeping with AHRQ's standards for employing a multidisciplinary approach to the development of systematic reviews, we called on our TEAG for inputs at two key points during this task. First, the group was asked to comment on the literature search and to give us feedback on our overall plans at that stage of the analysis, which included our approach to summarizing and synthesizing the evidence about cardiovascular outcomes associated with the use of epinephrine in the management of dental patients who are hypertensive. Second, they were asked to review the draft evidence report for completeness, correctness, and clarity.

Appendix C. Selection and Tasks of Peer Reviewers

Not including representatives of government agencies, we selected three organizations to contact to request that they suggest persons to serve as peer reviewers. In addition, the three members of the TEAG were asked to serve as peer reviewers because they played a major role throughout the project in conceptualizing the work and reviewing materials. We contacted all potential peer reviewers to determine their willingness to serve as peer reviewers, alerting them to the fact that this service would require them to prepare formal written reviews according to a checklist developed for this systematic review. Their comments and suggestions will form the basis of our revisions to the draft evidence report.

An important first step in the identification of potential peer reviewers for this systematic review was to determine the appropriate constituencies from which our reviewers should be drawn. We decided to include a representative of a dental education specialty professional organization involved with dental anesthesia (American Dental Education Association). We also chose to include a representative of patients and the public at large from the American Heart Association. Finally, we selected a representative from among dental trade associations and manufacturers (nominated by the Dental Manufacturers Association). The selection of representatives from government agencies rests with AHRQ and its National Institute of Dental and Craniofacial Research (NIDCR) collaborators. Clearly the sponsoring agencies as well as other government agencies providing dental services to a variety of populations are an important constituency, too. We believe that these categories represent a reasonable range of dental care experts, users, and patient groups to be involved in reviewing this report on the cardiovascular effects of epinephrine in hypertensive dental patients.