About the R Factor for Drug-Induced Liver Injury
Divide ALT by its upper limit of normal, divide ALP by its upper limit of normal, then divide the first by the second. An R of 5 or above is hepatocellular, 2 or below is cholestatic, and anything strictly between the two is mixed. The pattern is not cosmetic — it determines which of RUCAM's two scoring tables applies, and the two tables score the same case differently. Calculate it on the first abnormal set of liver tests, because the pattern drifts as the injury evolves and RUCAM expects the initial value.
Formula
R = (ALT ÷ ALT upper limit of normal) ÷ (ALP ÷ ALP upper limit of normal)
R ≥ 5 → hepatocellular
R ≤ 2 → cholestatic
2 < R < 5 → mixed- ALT ÷ ALT ULN
- ALT expressed as a multiple of the upper limit of normal. AST may substitute where ALT is unavailable.
- ALP ÷ ALP ULN
- ALP expressed as a multiple of the upper limit of normal.
- R is a ratio of ratios, which is why both upper limits of normal are required rather than the raw enzyme values alone. Using a remembered ULN can move the result across a threshold.
- Both thresholds are inclusive on their outer side: exactly 5 is hepatocellular, exactly 2 is cholestatic, and mixed is the strictly open interval between them.
- Calculate R on the FIRST abnormal set of liver tests. The pattern migrates as an injury evolves — many hepatocellular injuries become mixed and then cholestatic during recovery — and RUCAM is defined on the initial value.
- R has no upper bound and no lower bound other than zero. Very high values simply reflect a marked transaminitis with a normal ALP.
- The pattern says nothing about severity. Prognosis in a hepatocellular pattern is driven by bilirubin, through Hy's law, not by the size of R.
Interpreting the result
A hepatocellular pattern means hepatocyte injury predominates, and the immediate next step is to look at the bilirubin. Hy's law — hepatocellular injury with jaundice and no biliary obstruction — carries roughly 10% mortality and is the single most important prognostic signal in drug-induced liver injury; it is the reason the R factor is calculated early rather than retrospectively. A cholestatic pattern means impaired bile flow predominates, and the priority shifts to excluding mechanical obstruction with imaging before attributing anything to a drug. Cholestatic injury tends to resolve more slowly, sometimes over many months, but carries a lower short-term mortality. A mixed pattern is a genuine intermediate rather than an uncertain result, and is characteristic of a number of agents. In all three cases R is the input to a causality assessment, not the assessment itself — RUCAM, and increasingly RECAM, do that work.
| Score | Band | What it means | Action |
|---|---|---|---|
| R ≥ 5 | Hepatocellular | Hepatocyte injury predominates. Prognosis is driven by bilirubin — hepatocellular injury with jaundice and no obstruction is Hy's law, carrying roughly 10% mortality | Check bilirubin immediately; apply the hepatocellular RUCAM table |
| 2 < R < 5 | Mixed | Both mechanisms contribute. A genuine intermediate rather than an equivocal result, and characteristic of several agents | Apply the cholestatic/mixed RUCAM table; exclude obstruction as for cholestatic injury |
| R ≤ 2 | Cholestatic | Impaired bile flow predominates. Typically slower to resolve, sometimes over months, but with lower short-term mortality | Exclude biliary obstruction with imaging before attributing to a drug; apply the cholestatic/mixed RUCAM table |
What the R Factor needs (4 inputs)
- ALT (U/L)
- The measured alanine aminotransferase. AST may be substituted where ALT is not available, but it is a compromise — AST is not liver-specific and is raised by muscle injury and haemolysis.
- ALT upper limit of normal (U/L)
- Your own laboratory's ULN, which varies between assays and often by sex. This is the input most often filled in from memory, and getting it wrong shifts the pattern rather than merely the number.
- ALP (U/L)
- The measured alkaline phosphatase. Remember that ALP has bone, placental and intestinal isoenzymes — a raised ALP is not automatically hepatic.
- ALP upper limit of normal (U/L)
- Your laboratory's ULN for alkaline phosphatase.
What it returns
- R value
- A continuous ratio with no upper bound. Reported to two decimal places; the only thresholds are 2 and 5.
- Injury pattern
- Hepatocellular (R ≥ 5), cholestatic (R ≤ 2) or mixed (2 < R < 5).
- Which RUCAM table applies
- Hepatocellular, or cholestatic/mixed. RUCAM uses one table for hepatocellular injury and shares a second between cholestatic and mixed patterns.
How it is calculated
The R factor comes from the 1990 international consensus meeting convened to standardise how drug-induced liver disorders are defined and reported, work that also produced the definitions underlying RUCAM. The problem it solves is that raw enzyme values are not comparable across laboratories: assays differ, reference ranges differ, and an ALP of 300 means something different against a ULN of 120 than against a ULN of 280. Normalising each enzyme to its own upper limit of normal converts both to dimensionless multiples, and taking their ratio then expresses which of the two hepatic injury phenotypes predominates — hepatocyte necrosis, which releases transaminases, or impaired bile flow, which induces alkaline phosphatase. The cut-offs at 2 and 5 were set by consensus to leave a genuinely intermediate band rather than forcing a binary split, because a substantial proportion of drug injuries really do involve both mechanisms.
Facts & figures
| ALT | ALT ULN | ALP | ALP ULN | R | Pattern |
|---|---|---|---|---|---|
| 300 | 40 | 400 | 120 | 2.25 | Mixed |
| 300 | 55 | 400 | 105 | 1.43 | Cholestatic |
| 200 | 40 | 100 | 100 | 5.00 | Hepatocellular |
| 200 | 100 | 100 | 100 | 2.00 | Cholestatic |
The same enzyme values produce different patterns against different laboratory reference ranges — and a different pattern means a different RUCAM table. This is the practical reason the ULN is an input rather than a built-in constant.
| Criterion | Threshold |
|---|---|
| ALT alone | ≥ 5 × upper limit of normal |
| ALP alone | ≥ 2 × upper limit of normal (with GGT elevation, excluding bone source) |
| ALT with bilirubin | ALT ≥ 3 × ULN AND total bilirubin ≥ 2 × ULN |
These define when a biochemical abnormality is significant enough to be called drug-induced liver injury at all. They are separate from the R factor, which classifies the pattern once that threshold is crossed.
Evidence
International consensus meeting — Bénichou
1990The 1990 report of an international consensus meeting convened to standardise the definition and classification of drug-induced liver disorders, which established both the biochemical criteria for defining injury and the R ratio for classifying its pattern.
Provided the definitions on which subsequent causality assessment — including RUCAM — was built, and set the R thresholds of 2 and 5 by consensus.
RUCAM — the causality method that consumes the pattern
1993The Roussel Uclaf Causality Assessment Method, developed from the same consensus process, which scores causality using separate tables for hepatocellular and for cholestatic/mixed injury.
Operationalised the R-based classification: the pattern determines which table is used, and the two tables weight latency, dechallenge and risk factors differently, so the same case scores differently depending on the pattern assigned.
ACG clinical guideline
2021The 2021 ACG clinical guideline on the diagnosis and management of idiosyncratic drug-induced liver injury.
Retains the R-based classification into hepatocellular, cholestatic and mixed patterns as the standard framework, and sets out the biochemical thresholds defining DILI onset.
EASL clinical practice guidelines
2019The 2019 EASL clinical practice guidelines on drug-induced liver injury.
Uses the same R-based phenotyping and emphasises calculating it at the first abnormal liver test, since the pattern changes during the course of the injury.
How it compares
R Factor vs RUCAM
Sequential, not alternative — the R factor picks which RUCAM table to use, and RUCAM then scores causality.
RUCAM has two scoring tables, one for hepatocellular injury and one shared between cholestatic and mixed patterns, and they weight latency, dechallenge and risk factors differently. The same case scored on the wrong table produces a different causality category, which is why the R factor has to be settled first. The R factor carries no causality information of its own — it is purely a phenotype.
R Factor vs Hy's law
Complementary and sequential — the R factor establishes that the pattern is hepatocellular, and Hy's law then asks whether that hepatocellular injury is dangerous.
Hy's law identifies hepatocellular injury accompanied by jaundice in the absence of biliary obstruction, a combination carrying roughly 10% mortality. It cannot be applied until the pattern is known to be hepatocellular, which is what the R factor determines. Hy's law is prognostic where the R factor is purely descriptive.
R Factor vs Maddrey's discriminant function
Different diseases — Maddrey's grades severity in established alcohol-related hepatitis, while the R factor phenotypes a suspected drug injury.
Both involve abnormal liver biochemistry, but they answer opposite questions. Maddrey's assumes the diagnosis and grades it to guide steroid treatment; the R factor assumes nothing about cause and simply describes the biochemical pattern. Alcohol is in fact one of the competing causes RUCAM asks to be excluded, so the two frequently arise in the same differential without overlapping.
Pearls & pitfalls
- Enter your own laboratory's upper limits of normal, not remembered ones. R is a ratio of ratios, so a wrong ULN moves the pattern rather than just the number.
- Calculate R on the first abnormal liver tests. Patterns migrate during recovery — hepatocellular injuries commonly become mixed and then cholestatic — and RUCAM is defined on the initial value.
- Both thresholds are inclusive on the outer side. Exactly 5 is hepatocellular and exactly 2 is cholestatic; mixed is strictly between them.
- In a hepatocellular pattern, check the bilirubin immediately. Hy's law — hepatocellular injury with jaundice and no obstruction — carries roughly 10% mortality.
- In a cholestatic pattern, exclude mechanical obstruction with imaging before attributing anything to a drug. A stone in the common bile duct produces the same numbers.
- A raised ALP is not automatically hepatic. Check GGT or fractionate, since bone, placental and intestinal isoenzymes all contribute.
- The pattern is not a severity measure. A cholestatic injury is not milder than a hepatocellular one, and a very high R does not indicate a worse prognosis.
- Ask specifically about herbal and dietary supplements. They are a common cause and are very frequently not volunteered when patients are asked about 'medications'.
Critical actions
- Record the R factor at the first abnormal set of liver tests, with the laboratory reference ranges used, so it can be re-derived later.
- Where the pattern is hepatocellular, check bilirubin and INR at once and assess against Hy's law.
- Where the pattern is cholestatic or mixed, image the biliary tree before attributing the picture to a drug.
- Take a full exposure history covering prescribed drugs, over-the-counter agents, herbal remedies and dietary supplements, with dates of starting and stopping.
- Stop the suspected agent where the injury is significant, and document the dechallenge response — it is heavily weighted in RUCAM.
- Feed the pattern into RUCAM to score causality, rather than treating the R factor as a causality assessment in itself.
- Exclude the competing causes RUCAM asks about: viral hepatitis, autoimmune hepatitis, alcohol, ischaemia and biliary disease.
Why this score exists
The consensus meeting's real contribution was insisting on normalisation. Before it, drug hepatotoxicity was reported in raw enzyme values that could not be compared between centres, which made pooling cases across countries close to meaningless — and pooling is the only way to characterise reactions that occur in perhaps one in ten thousand exposures. Dividing each enzyme by its own laboratory's upper limit of normal makes a case from Lyon comparable with one from Osaka. The choice to leave a wide mixed band between 2 and 5, rather than splitting at a single point, reflects the same empirical caution: the committee were describing a spectrum of mechanisms and declined to impose a binary on it.
About the creator
Author of the 1990 international consensus meeting report
Reported the consensus criteria defining drug-induced liver disorders and the R ratio classifying their pattern.
Co-developer of RUCAM, from the same consensus process
Developed the causality assessment method that consumes the R-based pattern classification.
Limitations
- Purely descriptive. It classifies a biochemical pattern and carries no information about causality, severity or prognosis.
- Highly sensitive to the upper limits of normal entered, which vary between assays, laboratories and sexes — the same patient can be classified differently in two hospitals.
- The pattern drifts during the course of an injury, so a value calculated late in the illness will not match the one RUCAM expects and can misroute the causality assessment.
- Assumes the enzymes are hepatic in origin. A raised ALP from bone disease or a raised AST from muscle injury will distort the ratio without any liver pathology.
- The thresholds of 2 and 5 are consensus values from 1990 rather than empirically derived cut-points, and have not been re-validated against modern outcome data.
- Substituting AST for ALT, permitted where ALT is unavailable, degrades specificity because AST is not liver-specific.
- Says nothing about which drug is responsible when several were started together, which is often the hardest part of the real problem.
If you are the patient
When liver blood tests become abnormal and a medicine, herbal remedy or supplement might be responsible, doctors first work out what kind of liver injury it is. Two enzymes are compared: ALT, which rises when liver cells themselves are damaged, and ALP, which rises when the flow of bile is impaired. Each is measured against the normal range for that particular laboratory, and the two are then compared as a ratio called the R factor. A high value means the pattern is mostly liver-cell damage; a low value means it is mostly a bile-flow problem; values in between mean both. This matters for what happens next. A liver-cell pattern prompts an urgent check of the bilirubin level, because that combination can be serious. A bile-flow pattern prompts a scan to make sure there is no gallstone or blockage causing it instead. The R factor does not, on its own, prove that a medicine caused the problem — it decides which set of questions to ask next.
Frequently asked questions
What is the R factor in liver injury?#
The ratio (ALT ÷ ALT upper limit of normal) ÷ (ALP ÷ ALP upper limit of normal). It classifies the pattern of liver injury as hepatocellular (R ≥ 5), cholestatic (R ≤ 2) or mixed (between 2 and 5).
Why does the R factor need the upper limit of normal?#
Because raw enzyme values are not comparable across laboratories — assays and reference ranges differ. Normalising each enzyme to its own ULN converts both to dimensionless multiples, which is what makes the ratio meaningful. Using a remembered ULN can shift a case from one pattern to another.
What does a hepatocellular R factor mean?#
That hepatocyte injury predominates. The immediate next step is to check the bilirubin: hepatocellular injury with jaundice and no biliary obstruction is Hy's law, which carries roughly 10% mortality and is the most important prognostic signal in drug-induced liver injury.
What does a cholestatic R factor mean?#
That impaired bile flow predominates. Mechanical obstruction must be excluded with imaging before attributing it to a drug. Cholestatic injury tends to resolve more slowly than hepatocellular injury, sometimes over months, but carries a lower short-term mortality.
When should the R factor be calculated?#
At the first abnormal set of liver tests. The pattern migrates as the injury evolves — hepatocellular injuries commonly become mixed and then cholestatic during recovery — and RUCAM is defined on the initial value, so a late calculation can route the causality assessment to the wrong table.
Can AST be used instead of ALT?#
It can where ALT is unavailable, but it is a compromise. AST is not liver-specific and is raised by muscle injury and haemolysis, so substituting it reduces the specificity of the resulting pattern.
Does the R factor tell me whether a drug caused the injury?#
No. It classifies the biochemical pattern only, with no causality content. Causality assessment is done with RUCAM, and the R factor's job is to determine which of RUCAM's two scoring tables applies.
Is a cholestatic pattern less serious than a hepatocellular one?#
Not less serious, but different. Cholestatic injury typically takes longer to resolve while carrying a lower short-term mortality; hepatocellular injury can be rapidly dangerous when accompanied by jaundice. The pattern describes mechanism, not severity.
References
Original / primary reference
- Bénichou C. Criteria of drug-induced liver disorders. Report of an international consensus meeting. J Hepatol. 1990;11(2):272-276.
- Danan G, Bénichou C. Causality assessment of adverse reactions to drugs—I. A novel method based on the conclusions of international consensus meetings: application to drug-induced liver injuries. J Clin Epidemiol. 1993;46(11):1323-1330.
Guidelines
- Chalasani NP, Maddur H, Russo MW, Wong RJ, Reddy KR. ACG Clinical Guideline: Diagnosis and Management of Idiosyncratic Drug-Induced Liver Injury. Am J Gastroenterol. 2021;116(5):878-898.
- European Association for the Study of the Liver. EASL Clinical Practice Guidelines: Drug-induced liver injury. J Hepatol. 2019;70(6):1222-1261.