Onset and Resolution of Pain Among Treated and Untreated Posterior Teeth with A Visible Crack: Three-Year Findings from The National Dental Practice-Based Research Network

Abstract

OBJECTIVE:

Cracked teeth may be associated with pain, especially biting pain, and to a lesser degree cold and spontaneous pain. Described are how commonly these pains remain constant, develop, or resolve over time, none of which have been well-described, especially among untreated cracked teeth.

Methods:

Cracked teeth from the Cracked Tooth Registry (CTR) study were followed for 3 years. Assessments of cold, biting, and spontaneous pain and treatments performed were completed at enrollment (Y0) and at each annual recall visit.

Results:

209 practitioners enrolled 2,858 patients, each with a visible crack on a posterior tooth; 2,601 (91%) patients attended at least one recall visit. Overall, 960 (37%) were treated, primarily with crowns. Among both treated and untreated cracked teeth with biting pain or spontaneous pain at Y0, the vast majority (92–99%) had their pain resolved by the time of a recall visit and 85–93% remained pain-free after initial resolution. The observations for cold pain were similar: 68% (untreated) and 78% (treated) became free of cold pain at some point during follow-up, and 84% of these stayed free of cold pain after initial resolution. Few teeth developed biting or spontaneous pain (4–8%) and 44–67% of these had pain resolution during the follow-up period.

Conclusion:

In this study, treatment resolved a preponderance of pain associated with a cracked tooth. Pain was also resolved for most untreated cracked teeth, especially biting pain, and to a lesser degree spontaneous and cold pain, although not to the same degree as with the treated cracked teeth.

INTRODUCTION

Incomplete fractures, or cracks, in teeth are common, occurring in 80% of patients over age 40 [1]. The most reported symptoms of cracked teeth are pain to cold and on biting [2]. We previously reported types of pain present at enrollment in the Cracked Tooth Registry (CTR) study conducted as part of the National Dental Practice-Based Research Network (hereafter referred to as the Network) [3]. The CTR study enrolled a convenience sample of 2,858 patients, each with a posterior tooth with a visible crack; 45% had at least one type of pain. Although pain to cold was the most common type of pain, being present in 81% of the cracked teeth, in patients with any pain and in 37% of all the cracked teeth, cold pain alone was not associated with any tooth- or crack-level characteristic. Biting pain, although present in 35% of cracked teeth in patients with any pain and in 16% of all the cracked teeth, was associated with several tooth- and crack-level characteristics indicating the damage the crack had on the tooth. In the CTR study, cracked teeth with caries, biting pain, and spontaneous pain were much more likely to be recommended for restorative treatment, e.g., crowns, than those without these characteristics [4]. About 90% (389/435) of those treated at enrollment had an internal crack [5], supporting the treatment decision. In a CTR report of one-year follow-up [6] there was little progression of cracks. Among the 1,850 (65%) untreated teeth, only 6% developed an additional crack, while 32% of patients had a change in pain; pain decreases were twice as common as pain increases (23% vs. 10%).

Recently published were two reports on the entire three-years of follow-up of the CTR [7,8]. One paper focused on crack progression and development of fractures [7]. Of the 2,601 cracked teeth with any follow-up, only 3% (N=78) developed a fracture, and of the 1,889 patients untreated before year 1, 12% (N=232) had some type of crack progression. Ferracane et al [8] focused on treatments. Briefly, only 37 teeth were recommended to be extracted, 12 had only endodontic work (root canal therapy), 20 had endodontic treatment and restorations, and 925 had only restorations (the timing of procedures is available at Clinical Dental Study Results | The National Dental PBRN & Collaborative Group). Virtually all restorations were crowns, full or partial, some onlays. Compromised tooth structure, followed by sensitivity or pain, and broken or defective restorations (not mutually exclusive) were the most common reasons practitioners cited for recommending restoration.

There are several follow-up studies of cracked teeth; however, most are of treated cracked teeth focused on either restoration [9,10] or endodontics/root canal treatment [11,12]. As such, they do not report on pain, specifically, changes in pain. The only study that reported on pain in cracked teeth was a study conducted in a practice-based research network in the northwest U.S. In a three-year follow-up study of 670 randomly selected subjects and 153 non-randomly selected subjects conducted by the Northwest PRECEDENT network, in a mean follow-up of 1.5 years, almost 10% of previously pain-free cracked teeth developed some type of pain, and 25% of all cracked teeth showed some sign of crack progression [13]. The present report describes the changes in pain (resolution of pain; persistence of an existing pain; onset of a new pain) over three years of follow-up in the CTR study [14].

METHODS

A previous publication detailed the study procedures for the CTR study [14]. In brief, the study used a convenience sample of patients between 19–85 years old, each with at least one single vital posterior tooth in which at least one external crack was clinically observed, who were enrolled by dentists in the Network [15]. Dentist and practice personnel were trained in data collection; data were collected at the patient-, tooth-, and crack-level and all recommended and performed treatments and reasons for treatment. Data forms are publicly available at [http://nationaldentalpbrn.org/study-results/cracked-tooth-registry.php]. Tooth vitality was confirmed, primarily with cold [16] (e.g., refrigerant, ice), although some dentists used other methods such as air, air/water spray, or electric pulp testing. Spontaneous pain information was obtained via patient report, with pain to cold determined using the aforementioned methods, and pain on biting confirmed by having the patient occlude on a device or instrument placed on the occlusal surface of the cracked tooth. To help patients distinguish pain, i.e., a heightened response to the cold or bite assessment, from an ordinary response, dentists were asked to also perform these tests on a “normal” (e.g., contralateral) tooth. Participating dentists selected and characterized one eligible cracked tooth in each subject. Each practice enrolled up to 20 eligible subjects, or as many as they could in eight weeks, whichever came first. The Institutional Review Board (IRB) of the lead investigators (TH & JF), and the various IRBs that oversaw the six regions of the Network, reviewed and approved the study. All patients provided informed consent prior to participating.

Patients were requested to return to their dentist yearly for 3 years (Y1 – Y3). Efforts were made to schedule recall visits coincident with routine visits billed to the patient’s insurance. A 3-month window was permitted for the first two annual recall visits. Recall windows were opened for Y3, so that some patients had 24 months. Virtually the same patient-, tooth-, and crack-level information collected at enrollment (Y0) was collected at each yearly recall visit, including the presence of cold, biting, and spontaneous pain and treatments recommended and performed. All visits, other than Y0 and annual recall visits, during which the cracked tooth was treated, either planned or unplanned, were recorded as an interim visit. When a cracked tooth was extracted or referred for extraction, the tooth was withdrawn from the study; only 37 teeth were withdrawn for extraction. Overall, 209 practitioners enrolled 2,858 patients between 04/2014 and 04/2015; Y3 visit dates ranged 03/2017 to 12/2018.

Analysis.

Descriptive statistics, means with standard deviation (sd), and medians with inter-quartile range (IQR), were calculated for continuous variables. Changes in the presence of each type of pain were assessed separately for two groups: 1) cracked teeth treated surgically, namely, restorations (primarily crowns), endodontics (root canals), and extractions, at any time during the study (Y0 to Y3 inclusive); and 2) cracked teeth not treated, as the latter may represent natural changes or fluctuations in pain. When numbers were sufficient, associations of changes in pain with patient-, tooth-, and crack-level characteristics were assessed. Patients whose cracked tooth was recommended to be extracted, without attending any recall exam (N = 16), were excluded from analyses as only baseline pain assessments were available. Changes in pain among patients who attended a recall exam and whose (N = 21) cracked tooth was later recommended to be extracted were determined as any changes that occurred up to when recommended for extraction, and thus withdrawn. Significance of differences were adjusted for clustering of patients within practice using generalized estimating equations (GEE), implemented using PROC GENMOD in SAS with CORR=EXCH option. All analyses were performed using SAS software (SAS v9.4, SAS Institute Inc., Cary NC).

RESULTS

The mean age at enrollment in the Network of the 209 practitioners was 53 years (sd=10), median = 56 (IQR: 45 to 60). The majority were male (74%; N = 153), non-Hispanic white (84%; N = 173), and in a private practice (87%; N=181) (Supplemental Table 1). A total of 2601 (91%) patients attended at least one recall visit. Their mean age was 54 years (sd = 11.7), median = 55 (IQR: 46 to 62). The majority of patients enrolled were female (64%; N = 1,653), non-Hispanic white (85%; N = 2,190), had some dental insurance (78%; N = 2,017) and had a Bachelor’s degree or higher (53%; N = 1,365) (Supplemental Table 2). Of the 2,601 patients, 2,507 (96%) attended Y1, 2,236 (86%) attended Y2, 2,079 (80%) attended Y3, and 1,912 (74%) attended all three annual recall visits.

Surgically treated cracked teeth (Table 1)

Table 1.

Patients whose cracked tooth was treated at some time during the study, with changes in pain during the study by presence/absence of each type of pain at enrollment (Y0) (N=960).

			Type of pain present
Pain1 was present at Y0	Any type (N = 568)		Biting (N = 290)		Spontaneous (N=208)		Cold (N = 438)
Pain resolved, was absent at a recall visit	451	79%	277	96%	206	99%	340	78%
Pain stayed resolved or absent	424	94%	263	95%	195	95%	287	84%
Pain reversed to being present	27	6%	14	5%	11	5%	53	16%
			Type of pain absent
Pain was not present at Y0	No pain (N = 392)		Biting (N = 670)		Spontaneous (N = 752)		Cold (N = 522)
Pain developed, became present at a recall visit	102	26%	57	9%	43	6%	105	20%
Pain stayed present	50	49%	19	33%	14	33%	50	48 %
Pain reversed to being absent	52	51%	38	67%	29	67%	55	52%

¹278 (49% of 568 with any type of pain had 2 or more types of pain.)

Overall, 960 cracked teeth were treated surgically: 3 referred for extraction with no prior treatment, 12 treated only with endodontics, 20 with both endodontics and restoration, and 945 with restorations only, primarily crowns. Of the 960 patients whose cracked tooth was treated surgically during the study, most were treated prior to Y1. Specifically, 44% (N = 426) were treated at Y0 and an additional 30% (N=284) were treated after Y0 but prior to Y1, leaving 26% (N = 250) who were first treated at Y1 and later. Changes in pain by type and whether any change in pain, i.e., pain resolved or developed, persisted in the changed status, are presented in Table 1.

Virtually all the treated cracked teeth causing biting or spontaneous pain at Y0 became free of these pains: 96% (277/290) for biting pain, and 99% (206/208) for spontaneous pain. A total of 95% of each (263/277 biting; 195/206 spontaneous) stayed free of these pains throughout the study. For pain to cold, 78% (340/438) became pain free and 84% (287/340) of these remained free of cold pain.

Only 9% (57/670) of the patients whose cracked tooth was treated and did not have biting pain at Y0 developed biting pain during the study and in 33% (19/57) of these the pain persisted. Similarly, only 6% (43/752) developed spontaneous pain and in 33% (14/43) of these patients the pain persisted. Among patients whose cracked tooth was treated and who had no cold pain at Y0, 20% (105/522) developed cold pain at some time during follow-up, and in 48% (50/105) of these patients the pain persisted. The number of patients whose cracked tooth was treated and did not have biting or spontaneous pain at Y0 but did have pain after treatment was too small (57 for biting pain and 43 for spontaneous pain) to assess associations with any patient-, tooth- or crack-level characteristics, especially sustained increase or developed pain. The only characteristic associated with the onset of cold pain after having the cracked tooth treated was whether a crack connected with a restoration (23% [92/406] vs. 11% [13/116], P = 0.01).

Differences in pain before and after treatment are presented in the Figure. As shown on the left side of the Figure, among 710 patients whose cracked tooth was treated at Y0 or after Y0 but before Y1, the percent with any pain decreased from 64% to 27%, biting pain from 36% to 5%, spontaneous pain from 25% to 2%, and cold pain from 49% to 23%. All the reductions in pain among the 710 patients were significant at P<0.001. As shown on the right side of the Figure, among the 124 patients whose cracked tooth was first treated at Y1 or after Y1 but before Y2, the percent with any pain decreased from 40% to 19%, biting pain from 15% to 7%, spontaneous pain from 12% to 2%, and cold pain from 29% to 12%. All the reductions in pain among the 124 patients were significant at P<0.05. Decreases were of greater magnitude among the 710 patients treated earlier than the 124 treated later.

Figure:

Distribution of pain before and after treatment, among patients whose cracked tooth was treated between enrollment (Y0) and the recall visit at year 2 (Y2).

Cracked teeth not treated (Table 2)

Table 2.

Patients whose cracked tooth was not treated at any point during the study, with changes in pain during the study by presence/absence of each type of pain at enrollment (Y0) (N=1,641).

			Type of pain present
Pain1 was present at Y0	Any type (N = 586)		Biting (N = 116)		Spontaneous (N=113)		Cold (N = 507)
Pain resolved, was absent at a recall visit	396	58%	107	92%	107	95%	343	68%
Pain stayed resolved or absent	347	88%	91	85%	100	93%	288	84%
Pain reversed to being present	49	12%	16	15%	7	7%	55	16%
			Type of pain absent
Pain was not present at Y0	No pain (N = 1055)		Biting (N = 1518)		Spontaneous (N = 1521)		Cold (N = 1134)
Pain developed, became present at a recall visit	226	21%	75	5%	56	4%	214	19%
Pain stayed present	123	54%	42	56%	27	48%	108	50%
Pain reversed to being absent	103	46%	33	44%	29	52%	106	50%

¹117 (20% of 586 with any type of pain had 2 or more types of pain.)

Decreases in pain were also observed among the 1,641 patients whose cracked tooth was not treated at any time during the study and who had biting pain (N=116) or spontaneous pain (N=113) at Y0, 92% and 95% (N=107 for each), respectively, specifically, patients no longer had biting pain or spontaneous pain after Y0. Furthermore, the vast majority of each, 85% (N=91) for biting pain and 93% (N=100) for spontaneous pain, did not revert to having pain; patients stayed pain-free from biting pain and spontaneous pain for the remainder of the study. The pattern for cold pain was like biting pain and spontaneous pain, although not as strong. Cold pain was resolved in only 68% (343/507) (compared to 92% and 95%) of those who had cold pain at Y0; namely, they no longer had cold pain at some time during the study, and most (84%; N=288) stayed free of cold pain for the duration of the study.

Among patients with cracked teeth who were not treated and who presented without biting pain or spontaneous pain at Y0, only 4% (56/1521) without spontaneous pain and 5% (75/1518) without biting pain developed biting pain or spontaneous pain, respectively. Thus, the vast majority (95–96%) of cracked teeth that were not treated and did not cause biting or spontaneous pain at Y0, remained free of these types of pain for the three years of follow-up. About half of the few that did develop biting pain (44%, 33/75) or spontaneous pain (52%, 29/56) reverted to being free of these pains before the end of the study. Of the non-treated cracked teeth that did not have cold pain at Y0, 19% (214/1134) developed cold pain at some time during the three years of follow-up; thus 81% remained free of cold pain. This was similar to, but not as high as, the 95–96% that remained free of biting pain and spontaneous pain for the duration of the study. As with the latter types of pain, half of those who did develop pain to cold (50%, 108/214) reverted to being free of cold pain before the end of the study.

Changes in cold pain by patient-, tooth-, and crack-level characteristics.

The distributions of patients are presented in Supplemental Tables 3–5 by patient-, tooth-, and crack-level characteristics and change in cold pain status. Independent associations of these characteristics with changes in cold pain are presented in Supplemental Table 6. Changes in biting and spontaneous pain were too small to permit statistical analysis. Among cracked teeth for which cold pain was present at Y0, the cold pain was more likely to resolve, i.e., not present, at a recall visit, for patients from the Midwest region, and for males, and older persons, while cracks on teeth with roots exposed were less likely to have cold pain resolve. Findings were similar for a ‘consistent’ decrease, specifically, excluding 108 patients who reverted and presented with cold pain at a subsequent recall visit.

Among cracked teeth for which cold pain was not present at Y0, the cold pain was more likely to develop, i.e., be present at a recall visit, for cracks on molars and less likely in patients from the Midwest region. Findings were similar for „consistent’ increase (excluding 161 patients whose cold pain reverted to being absent after developing), with the added characteristic of age, namely, cracks among persons younger than 55 years of age were more likely to have cold pain develop and stay present at subsequent recall visits than persons 55 years of age and older.

DISCUSSION

Among both treated and untreated posterior cracked teeth with biting pain or spontaneous pain at Y0, the vast majority, 92–99%, became free of these pains and 85–95% stayed free of these pains. Cold pain was similar, but not to the same degree: 68% (untreated) and 78% (treated) of patients who had cold pain at Y0 became free of cold pain during follow-up and 84% of these stayed free of cold pain. Few patients developed biting or spontaneous pain (only 4–9%) and only 33–56% of those patients who developed biting or spontaneous pain sustained these changes. Only 18–20% of patients developed cold pain and 48–80% of patients sustained this change.

Among treated cracked teeth, the timing of reduction in pain relative to when treatment was performed supports the explanation that the treatment was successful in reducing or eliminating pain, especially biting pain and spontaneous pain. Presence of pain was a common reason dentists stated for recommending cracked teeth be treated [4]. Another study that reported on change in symptoms/pain in a follow-up study of cracked teeth [13] reported increases only; 10% of previously pain free cracked teeth developed some type of pain in a mean follow-up of 1.5 years of 823 patients. They did not report the type of pain or whether there was any decrease or resolution of pain. The finding in our study that 21% of patients whose cracked tooth did not cause pain at Y0 did ultimately report pain at some time in 3 years (primarily pain to cold), is like the 2012 report from Hilton and colleagues. In the present study, however, about half of patients who developed pain reverted to being pain-free by the end of the study.

When the CTR study was initially proposed, an underlying hypothesis was that most pain present at Y0 would persist through the follow-up period, resulting from crack progression or its manifestation. This hypothesis is not supported by the findings reported here in which the majority all types of pain present, either at Y0 or later, resolved for a large percentage of teeth. Numbers were only sufficient to assess associations with changes in cold pain.

Possible reasons for the decrease in pain can be speculated, e.g., people become tolerant of pain, cracks produce leakage that stimulates secondary dentin to form and reduce dentinal tubule fluid flow causing pain, or cracks fill with debris that does the same. To our knowledge there are no longitudinal studies in the literature to support or refute these speculations. Another possible reason, primarily for pain to cold, is that some of the reports of pain due to cold stimuli could reflect dentin hypersensitivity rather than tooth damage by the visible crack. Dentin hypersensitivity is common [17,18,19], cold stimulation is the most common stimulus patients report for it, and it can be treated with over-the-counter products such as desensitizing potassium nitrate toothpaste [20,21]. As previously reported from the CTR study at Y0 [3], cold pain alone was not associated with any tooth- or crack-level characteristic, which would be consistent with at least some cold pain being attributable to a cause other than tooth damage associated with a crack. A limitation of the current study is that the practitioners, patients, and cracked teeth were not selected randomly. There was continuous enrollment, however, and all patients who qualified and consented were enrolled in the study, which should have minimized any potential selection bias. Also, each type of pain was assessed dichotomously (present/absent), rather than on a continuous numerical scale, so changes in the severity of pain could not be assessed. The inability to assess whether there was any patient-, tooth- or crack-level characteristics associated with changes in biting pain or spontaneous pain, especially decreases in biting pain among patients whose cracked tooth was not treated, was also a limitation. Biting pain is the type of pain most closely associated with having a cracked tooth [2]. As we started with 2,885 cracked teeth, 1,641 (57%) had not been treated after 3 years, and only 116 (7%) had biting pain at Y0, of which 95% (N= 107) became free of biting pain at one or more recall visits at Y0. Thus, numbers for comparison would involve 107 to 9, too small for any meaningful comparison. If the same proportions held, it would be necessary to enroll 50,000 teeth to have 100 for comparisons of cracked teeth that retained biting pain for 3 years. Such a study would not be feasible to conduct.

The study has several strengths. The first is the size, namely, the large number of practitioners and cracked teeth included. Another strength is the geographic diversity of the study population, which was from all regions of the continental United States, and the diversity of practice settings involved, from small solo private practices to large, preferred provider organizations. Lastly, changes in types of pain were assessed separately for treated and not-treated cracked teeth.

CONCLUSION

This is the first study to report on changes in pain from cracked teeth in a large study over multiple years. Most pain resolved over the course of the study, especially among teeth that were surgically treated with a restoration. The study demonstrates that few of the pains from untreated posterior cracked teeth are constant, and most resolve over time.

CLINICAL SIGNIFICANCE.

Restorative treatment resolves virtually all biting and spontaneous pain and a preponderance of cold pain associated with a posterior cracked tooth. Few of the pains from untreated posterior cracked teeth are constant, most resolve over time.