The Probation Service has said it has begun a review of a case in which errors were contained in a report about a sex offender who was jailed earlier this week for 18 and a half years.

However the director of the service, Vivian Geiran, said he stood over the revised report which placed the offender at a medium risk of re-offending.

On Thursday a judge at the Circuit Criminal Court said she found it difficult to accept a probation officer's assessment of the man's risk of re-offending.

The judge said the serial attacker was "clearly a danger to society and to women" and had used extreme violence of a "brutal and primal nature" against three women.

Judge Pauline Codd was deciding on the sentence of a man who had violently attacked three women in the Clondalkin area between 2011 and 2016. Two of the women had been sexually assaulted.

A probation officer had assessed him of being at medium risk of reassessment - the second lowest point on a four point scale.

The judge had been due to sentence the man on Monday but when a probation report that she had ordered was handed in to court it contained a number of errors.

The wrong name was used a number of times in the report and there was a reference to previous convictions for road traffic offences, when in fact the defendant in the case had none.

Prosecuting counsel said there were concerns about the report and it was their duty to bring these concerns to the attention of the judge.

The judge adjourned sentence until Thursday and asked for a representative from the Probation Service to attend court.

The probation officer who had written the report attended court on Thursday with a new report and described the errors in the previous report as "an oversight".

When pressed by the judge as to what she meant by "oversight" she said she had been working on a number of reports at the same time and may have entered the wrong name by mistake.

She said she had spent two hours with the defendant and an interpreter but did not deem him suitable for probation because he would not engage with her.

The judge asked if she could stand over the risk assessment in the report which put the defendant at medium risk of re-offending. She said she could.

The judge asked how they had reached that conclusion about a man who had attacked three women in separate attacks over a number of years and who had refused to engage in any discussion with the probation officer about his behaviour, had not taken responsibility for his actions and had shown no remorse.

The probation officer replied that it was "an actuarial assessment" and for the purposes of a sentence hearing the offences were counted as one because it was one sentence hearing.

She said there could be multiple offences but because he had no previous convictions and this was the first sentence hearing it was counted as one for the purposes of assessing the risk of re-offending.

Judge Codd said she could not accept that the man was at medium risk of re-offending. She gave a detailed sentence structure to include consecutive sentences totaling 18 and a half years.

She also ordered that he complete a sex offenders programme while in prison and be supervised for nine years after his release or risk being returned to jail.

On RTÉ's This Week programme Mr Guerin accepted there were factual errors in the original report but said these had been corrected quickly.

He said a full review would be carried out to establish how the errors occurred and to help improve practices where necessary.

However he said placing someone as being at medium risk of re-offending was still a serious categorisation and the methods used by the Probation Service were used in many other jurisdictions.

The RM 2000 was an initial screening tool which was widely used, he said. The probation service would continue to work with the offender while in custody and post release, he said.

"The fact that someone is (assessed as) medium risk is in our view very serious. We will in this case continue to work with the individual, " he said adding that this was an indication "how serious we take this and the level of effort we are prepared to put into individual to try ensure they do not reoffend."

Mr Guerin said "if there are particular lessons to be learned from this we will learn and make whatever changes we need to make for the future."

He said the service was "committed to providing quality services of the highest standard and it is important they are trusted by the courts and the community."

In response to written questions from the programme the service also said:

"The Probation Service operates a dual system of risk assessment when working with sexual offenders - the RM2000 and Stable and Acute 2007 (S&A2007). The initial assessment involves the application of the RM2000. The RM2000 identifies the likelihood of a reconviction for a sexual offence based on facts about past personal and offending history. 

 "The RM2000 does not indicate which factors need to be addressed in supervision/ treatment, for risk to be reduced. It is therefore normally used as an initial assessment and screening instrument to inform which offenders warrant further assessment.

In completing a further assessment, the S&A2007 is then applied.  The S&A2007 measures dynamic risk factors associated with each individual offender.  These identified risk factors help the Probation Service to identify treatment/supervision targets and form the basis of a case management plan for the supervision period.  The S&A 2007 is normally completed at assessment stage or commencement of a period of supervision.

The RM2000 was applied in this case concerned.

The Risk Matrix 2000 (RM2000) is an actuarial risk assessment instrument designed to identify sexual and non-sexual violent reconviction among men who have been convicted of a sexual offence. 

It classifies an offender into one of four risk groups low, medium, high and very high based on some simple facts about his criminal and personal history.

Each of these risk categories differ in their rates of re-conviction for sexual offences.  The low risk group have a lower re-conviction rate than the medium risk group and so on. 

The RM2000 cannot tell us who will or will not offend, it can however tell us whether the offender has similar characteristics to a group of offenders who have a higher or lower rate of known re-conviction.

The RM2000 is the first important step in assessing risk and is used by the Probation Service and An Garda Síochána as a baseline measure to assess the risk posed by sex offenders referred to either or both agencies.

Asked why the multiple offences can be treated as one, the service said:

"In applying the RM2000, the assessing Officer is required to count the number of times the offender appeared in court convicted and sentenced, in relation to one or more sexual offences. In instances where the offender is appearing before the Court on the same day for multiple matters, this appearance is counted as one 'sexual sentencing appearance' for the purpose of the RM 2000 - the RM2000 does not focus on the number of sexual offences committed. The underlying concept is whether an offender has persisted to offend after the last time he was punished (i.e convicted and sentenced by a Judge).