Last Friday, having reserved the decision to himself, the Director of Public Prosecutions announced that, in this 'unique case', Harold Shipman will not face further murder charges.This despite the fact that the police have stated that there is sufficient evidence to charge the former general practitioner with 23 counts of murder in addition to the 15 on which he was convicted last month.

Furthermore, most of the relatives of his other suspected victims have stated that they favour further prosecutions.So why has the DPP decided as he has done? In resolving whether to prosecute Shipman again, the DPP needed to consider a large number of legal and public interest factors.

Perhaps the most critical was whether, in the light of the massive publicity surrounding Shipman's trial and conviction, any judge called to preside over a future trial would allow it to proceed.

Would instead such a judge hold that, because of the publicity, Shipman could not receive a fair trial and therefore dismiss the prosecution before it ever started? If this happened, the Crown Prosecution Service would look stupid for having bothered to prosecute an untriable case; worse still, from the relatives' perspective, Shipman would be entitled to be acquitted in relation to all counts of murder for which he was to have been tried.The law on pre-trial publicity is an aspect of a body of jurisprudence labelled 'abuse of process'; a power vested in any criminal court to prevent a prosecution of a defendant on the ground either that he could not receive a fair trial or, for some policy reason, it would be improper to allow the trial to proceed.

A judge has an inescapable responsibility to ensure that any trial is fair.

In the context of a jury trial, fairness must mean that it can safely be presumed that the 12 persons called upon to try the defendant will do so dispassionately and only on the evidence they hear in court.Having for generations stoutly defended the principle that trial by jury is always fair, in recent years judges have conceded that pre-trial publicity can cause a risk of prejudice.

Trials have been stopped particularly because of some past criminal history of a defendant -- a fact previously hidden from the jury -- being publicised.Applying this law to the Shipman case, it seemed highly likely that a judge would thwart any further prosecution of him for murder.

Several factors are especially pertinent: it would be impossible to empanel a jury which would not already know of Shipman's convictions.

Moreover, the evidence supporting any fresh indictment would be similar to the extensively reported evidence that led to those convictions.In addition, Shipman's 1970s conviction for drug abuse has been revealed.

If there is doubt as to the potentially prejudicial effect of this conviction, the attitude of the trial judge, Forbes J, should be recalled.

When he learnt that the BMA had circulated news of this conviction to its members while the jury was still out, he denounced this as capable of prejudicing the fairness of the trial.

He went even further and opined that had the jury learnt of this, the entire trial would have had to have been abandoned.

To his mind a 'flagrant contempt' had been committed.If this was the view of the trial judge about what is now a relatively insignificant piece of 1970s information, surely in the light of the massive post-conviction publicity all of it condemnatory of Shipman, the case for a stay in relation to any future trial was unanswerable.

How could another judge properly ignore the firm opinion of Forbes J?If, as is now the case, the criminal justice system is unable to satisfy the legitimate demands of the relatives for the truth to be uncovered via a retrial, how can their wishes be met? It can probably only be achieved by a full public inquiry charged with the task of investigating all relevant deaths will suffice.

Leaving this to a coroner alone will not be adequate.