A gingival retraction cord looks like the simplest item on the tray, yet the choice of cord, medicament and timing decides whether the margin is readable and whether the next step, impression, scan or bonding, works as intended. This article focuses on the decisions around the cord: which construction, size and medicament to use, how long to leave it, when a PTFE cord is the better first cord, and what haemostatic agents do to impression materials, scans and adhesive cementation. Every figure comes from Dr Hajtó's protocol, a study or manufacturer information.
What the cord has to achieve
Retraction has two jobs: displace the gingiva far enough for the margin and some tooth structure beyond it to be captured, and keep the sulcus free of blood and crevicular fluid while you record it. For elastomeric impressions there is a useful benchmark. In a model study, Laufer and colleagues found defects in 70 to 100 % of impressions when the simulated sulcus was only 0.05 mm wide, inconsistent results at 0.10 and 0.16 mm, and little difference in distortion between 0.2 and 0.4 mm. A sulcus width of about 0.2 mm is therefore a sensible target, and it is the width Baharav and colleagues used as their benchmark.
Intraoral scanners are less forgiving, because they can only record what the light reaches. A systematic review of seven studies by Andrade Villalobos and colleagues found that scanner trueness is significantly compromised at subgingival margins, that cord retraction improved accuracy and that saliva contamination reduced it. Clinically acceptable results depended on proper retraction and a dry field.
Retraction does not compensate for a margin that is placed too deep. If the finish line invades the attachment, no cord will make it readable without trauma; the background is explained in the article on biologic width.
Choosing a gingival retraction cord
Construction and size
Cords are knitted, braided or twisted, and come plain or impregnated in several diameters. Kumbuloglu and colleagues compared 16 cord systems clinically in a blinded design. Impressions were better with cords that were easy to insert, dilated the sulcus properly, caused little recurrent bleeding and left no remnants or loose fibres. Both cord thickness and medicament were significant risk factors for the quality of the gingival margin area. Untreated, medium-braided and epinephrine-impregnated systems were the clinically successful ones in that study.
In practice that means:
- First cord: thin enough to sit below the finish line without covering it.
- Second cord: thick enough to displace the tissue laterally, but not so thick that it has to be forced in.
- No fraying: loose fibres at the margin end up in the impression, the scan or the luting resin.
Medicaments compared
| Cord or agent | Effect | Evidence and cautions |
|---|---|---|
| Plain cord | purely mechanical | no chemical side effects; relies on local anaesthetic and technique for haemostasis |
| Aluminium chloride | astringent | in a randomised trial, 25 % aluminium chloride cords left for 20 minutes caused 88 ± 4 µm of recession after 28 days; limit to 10 minutes (Fouad et al. 2025) |
| Ferric sulfate / ferric subsulfate | strong haemostasis | Ultradent describes Astringedent X as a 12.7 % iron solution; the manufacturer warns against mixing it with epinephrine or using it with epinephrine-impregnated cords (blue discolouration of the soft tissue); residue impairs polyether, so rinse thoroughly; one case report links ferric sulfate to dentine discolouration under crowns |
| Epinephrine | vasoconstriction | reduced gingival blood flow and a fall in diastolic pressure in healthy volunteers; considered safe in healthy gingiva when stress and tissue trauma are avoided (Polat et al. 2007); do not combine with ferric sulfate agents (manufacturer warning) |
For epinephrine cords, systemic effects deserve attention. Kumbuloglu and colleagues explicitly call for considering them, and a systematic review by Bader and colleagues found that the available studies did not address the effect of epinephrine retraction cords on hypertensive patients at all. Where systemic epinephrine is a concern, a cord without epinephrine avoids the question.
The double-cord technique: Dr Hajtó's protocol
The double-cord technique is the workhorse for subgingival margins. Dr Hajtó combines it with local haemostasis and, where needed, the electrotome:
- Anaesthesia with papillary infiltration: after the regular anaesthetic, infiltrate small amounts of an articaine solution with 1:100,000 adrenaline around the tooth and into the papillae, blow air over the site, and infiltrate a second time. The vasoconstrictor reduces bleeding from the papillae during preparation and cord placement.
- First cord: after the initial preparation, place a thin first cord circumferentially. A bimanual technique helps: one instrument holds the section already placed while the second packs the next one.
- Finish with the cord in place: in Dr Hajtó's protocol, finishing with the first cord in place keeps the finish line out of the supracrestal attachment.
- Second cord: pack a thicker cord on top and leave it for 6 to 8 minutes. This is longer than the 4 minutes that Baharav and colleagues found optimal for a 0.2 mm crevice (2 minutes gave significantly narrower crevices, 6 and 8 minutes no further gain), and below the 10-minute limit for aluminium chloride cords.
- Remove the second cord: moisten the cord, withdraw it slowly and rinse with air and water spray. Haemostatic agents for residual bleeding, and the cautions that go with them, are covered in the medicament table above.
- Record immediately: impress or scan straight away with the first cord in place. In Baharav's study, the sulcus at the transitional line angles was narrower than mid-buccally for up to 3 minutes after cord removal, so start where it closes fastest.
- Check and correct: inspect the scan. If tissue still covers the margin at individual sites, trim it there with the thinnest wire electrode while the first cord is in situ, then re-scan.
Dr Hajtó uses the electrotome only interdentally, and in the anterior region only to remove excess tissue. On the facial aspect he applies it with particular caution, and always with the thinnest electrode.
Hands-on dental courses
Courses for dentists in Munich, several of them taught in English.
See all coursesPTFE cord as the first cord
As an alternative to the conventional cord, Dr Hajtó uses polytetrafluoroethylene (PTFE) cord. He uses diameters of 0.6, 1.0 and 1.8 mm, and also twisted PTFE tape. The material behaves differently from textile cords:
- One cord is often enough: PTFE is volume-stable and keeps the sulcus open instead of collapsing.
- No fibres, no soaking: it does not fray and does not absorb blood or fluid, so the margin stays clean and dry.
- No bonding: its smooth surface does not stick to provisional resin, adhesives, impression materials or the packing instrument.
- Easy to see and remove: it is unlikely to be left behind or trapped during cementation.
- Scanner-friendly: it can stay in the sulcus during intraoral scanning.
The concept has early clinical support. In a pilot case series by Ruggiero and colleagues, a 0.076 mm PTFE tape was left in the sulcus of 20 teeth with subgingival margins during scanning. The finish line was fully visible in 90 % of scans, a sulcus expansion of at least 0.2 mm was achieved in 90 % of teeth, and there was no bleeding or laceration.
Because PTFE does not bond to resin, it is also useful during adhesive cementation of veneers and crowns with margins close to the gingiva.
What haemostatic agents do to the next step
Impression materials
The fear that haemostatic agents inhibit addition silicones is not supported by the available data. Machado and Guedes tested ferric sulfate, aluminium sulfate and aluminium chloride against three polyvinyl siloxanes, as well as cords handled with latex gloves, in 240 samples; none of the combinations affected setting. In Kumbuloglu's clinical comparison, none of the cord medicaments affected the polyvinyl siloxane surface either. Polyether is a different matter: the manufacturer of Astringedent X states that residual ferric sulfate on the preparation negatively affects polyether, and recommends washing the area completely.
Adhesive cementation
For bonded restorations, contamination matters more. A systematic review by Bourgi and colleagues found that any contamination of dentine with blood or saliva impaired bond strength. Water spray and reapplication of the bonding system could reverse the effect of blood or saliva, but using haemostatic agents as a way of decontaminating blood was not recommended. If haemostatic agents were needed, keep them away from dentine that is about to be bonded and clean the surface. For Astringedent X, Ultradent specifies a brief 2 to 3 second cleanse with 35 % phosphoric acid followed by a firm air/water spray before bonding or impressions; with an etch-and-rinse bonding system the regular etching step covers this. The material choice also determines how dry the field has to be; the article e.max vs zirconia compares cementation options.
Discolouration
Conrad and Holtan described a patient whose incisors showed internal dentine discolouration under porcelain crowns four years after placement. They proposed ferric sulfate combined with a knitted cord as a possible cause, through precipitation of iron compounds in demineralised dentine. This is a single clinical report and a hypothesis, but it is another reason to rinse iron-based agents off completely.
Troubleshooting
| Problem | Likely cause | Fix |
|---|---|---|
| Margin hidden on the scan | sulcus closed after cord removal, or first cord too thick | scan immediately, start at line angles; use a thinner first cord or PTFE |
| Fibres in the impression | frayed cord at the margin | change cord type; use PTFE as first cord |
| Bleeding after cord removal | inflamed gingiva, cord pulled dry | moisten before removal; ferric sulfate haemostatic on its own or with a plain cord (never with epinephrine), then rinse |
| Recession weeks later | aluminium chloride cord left too long, trauma during packing | keep aluminium chloride to 10 minutes or less; pack obliquely, not with force |
| Tissue covers margin locally | excess or hyperplastic tissue | fine wire electrode interdentally with the first cord in place, then re-scan |
| Inhibited polyether surface | residual ferric sulfate | rinse thoroughly before the impression |
Summary
A gingival retraction cord works best as part of a sequence: papillary infiltration for haemostasis, a thin first cord that stays in place while you finish and record, a second cord for 6 to 8 minutes, careful removal and immediate recording. Choose the medicament with the patient and the next step in mind, keep aluminium chloride under 10 minutes, rinse iron-based agents thoroughly, and consider PTFE where fibres or bonding would be a problem. Tissue management, scanning and cementation are practised hands-on in the prosthodontics courses at Munich Dental Academy.
Frequently asked questions
How long should a gingival retraction cord stay in place?
Baharav and colleagues found 4 minutes to be the optimum for a crevice of about 0.2 mm after removal; 2 minutes gave narrower crevices, and 6 or 8 minutes gave no further gain. Dr Hajtó leaves the second cord for 6 to 8 minutes. With aluminium chloride cords, 10 minutes should not be exceeded, because 20 minutes caused measurable recession in a randomised trial.
Which size retraction cord should I use?
Use a thin first cord that sits below the finish line without covering it, and a thicker second cord that displaces the tissue laterally without being forced in. In a clinical comparison of 16 cord systems, cord thickness was a significant factor for the quality of the gingival margin area. PTFE cord, which Dr Hajtó uses in 0.6, 1.0 and 1.8 mm, often makes a second cord unnecessary.
Are epinephrine retraction cords safe?
In healthy volunteers, epinephrine cords reduced gingival blood flow and were considered safe in healthy gingiva when stress and tissue trauma are avoided. However, potential systemic effects must be considered, and a systematic review found no studies on epinephrine cords in hypertensive patients. Where systemic epinephrine is a concern, use a cord without it. Never combine epinephrine cords with ferric sulfate haemostatics.
Do I need a retraction cord for an intraoral scan?
For supragingival margins usually not. As soon as the margin is equigingival or subgingival, the scanner needs a clear, dry view. A systematic review found that scanner accuracy at subgingival margins was compromised without retraction and improved with cord retraction, while saliva made it worse. Leaving the first cord or a PTFE cord in place during the scan is the most reliable approach.
Do haemostatic agents affect impressions or bonding?
In a laboratory study, ferric sulfate, aluminium sulfate and aluminium chloride did not inhibit the setting of polyvinyl siloxanes. Residual ferric sulfate does affect polyether, according to the manufacturer, so rinse thoroughly. For bonding, any blood or saliva contamination lowers dentine bond strength, and haemostatic agents are not recommended for decontamination, so keep them away from dentine you are about to bond.
