PSA testing and a missed diagnosis

By Dr Bryan Betty, College Medical Director and Dr Peter Moodie, College Clinical Advisor

14 April 2022

Category: Clinical



The more HDC reports you read, the more you see common themes developing and, in this case, they include:

  • A classic 鈥淪wiss cheese鈥 where a series of small errors created a large error.
  • A guideline (MoH Prostatic Specific Antigen (PSA) testing) was not followed and the guideline itself was confusing.聽 The guideline was in fact considered a protocol.
  • A practice struggling with understaffing, at least partly due to doctors retiring and having to be replaced by locums.
  • A results management policy which with the best of intentions involved numerous people ordering, checking, and acting on test results.
  • Inconsistent or incorrect warnings on laboratory reports.

The case

The case involves a man in his 60s who knew he had a family history of prostate cancer, which he stated at every opportunity to his doctor. His medical history was complicated and on his first visit to the medical centre he was reviewed and managed very appropriately (including having 10 medications prescribed). His family history of prostate cancer was not recorded but he was given a form for a PSA test.

In context, the non-recording of the family history may well have been because the consultation was long and thorough. The error could have been corrected by asking the patient to return for a prostate review (assuming the patient did tell the doctor about the family history).聽 In fact, the PSA result was moderately, but not excessively, abnormal at 5.18ng/ml. The MoH guidelines state that for this man鈥檚 age the test should be repeated in three months and if still raised, the patient should be referred. The purpose of the three-months repeat of the test was to exclude false positives, not to identify a rising titre. When the doctor saw the PSA result, he inexplicably recorded 鈥渘o red flags鈥 in the comments section and asked for a repeat in six months鈥 time.聽 聽

From here, things went wrong; the doctor left the practice, but he did hand over the case to another doctor who had never seen the patient, so when the second test came back (still slightly raised) the patient鈥檚 new doctor relied on the earlier comment of 鈥渘o red flags鈥. A further test was ordered by a prescribing pharmacist who followed a practice policy of assigning screening test results, ordered by someone like a pharmacist or nurse, to the name of a permanent doctor. It appears that, in this case, a random doctor鈥檚 name was used.

The practice had a complicated laboratory test review and management policy which was designed to minimise errors; however the reality was that it involved a number of players. The policy had been designed because of the short staffing and the reliance on locum doctors.聽

Without going into detail, the patient was finally referred to the urology department some 25 months after the initial consultation, and with a with a PSA of 6.32. A biopsy showed that he did have prostate cancer.

The Commissioner was critical of the doctors involved for their errors, and for not following guidelines. The Commissioner also criticised the practice management policy and asked why there was no specific policy for the follow up of abnormal PSA tests. The College was consulted on this and informed the Commissioner that although it was a Cornerstone requirement to have a results management policy, it was not necessary or indeed practical, to have it cover specific tests.


Clinical errors did occur with this patient; however, there are other learnings:

  • The Commissioner regards 鈥済uidelines鈥 as 鈥減rotocols鈥 and therefore you need to have good reasons for ignoring them. This can be worrying if the guideline is a consensus document (often the lowest form of evidence). In this case there is some confusion about how the guideline should be implemented and the Commissioner acknowledged this. There is a curious anomaly where a 70-year-old man鈥檚 abnormal result is anything above 4ng/ml while a 71-year old鈥檚 abnormal level is above 10ng/ml. For me, this suggests judgement calls rather than relying on careful data. Further, the need for a three month repeat of an abnormal test is to see if it has reduced, not to see if it has increased.
  • There is a worrying trend for laboratories to be inconsistent in the warnings and advice they give out. This has been highlighted in at least one other report and needs to be focused on by the Commission.聽 '
  • Cornerstone requirements were heavily relied on by the Commissioner and indeed these protected the practice from overly onerous protocols. Practice management is coming under increased scrutiny and trying to say 鈥渋t is up to the doctor鈥檚 discretion鈥 is not going to wash anymore.聽 Corporate responsibility is now a reality.
  • Management of abnormal test results is again a continuing theme. The processes should be as simple as possible and using portals like MMH can be crucial. By using a portal, the patient receives the result and advice directly from the doctor and can check the actual report themselves. This makes it even more important for lab tests to have accurate advice attached.
  • Doctor and staff shortages are not going to be an excuse for errors in care. With 50 percent of the general practice workforce retiring by 2030 this will be a real issue.聽 '
  • With the increasing use of multidisciplinary teams to cover these shortages there is a real danger of ending up with a 鈥渓eaderless clinical group鈥. This was highlighted in a previous article on mental health services in DHBs and general practice needs to make sure it does not fall into the same trap.
  • As general practitioners managing things like screening, we tend to think in three, six or 12 monthly cycles. When there is no continuity of care there is a risk that things can 鈥渄rag on鈥 without us noticing it; in this case it was 25 months.

Lastly, despite this man being at high risk of malignancy I was dumfounded to see that the DHB urology service was not criticised for taking five months from the referral to carrying out a biopsy.聽

Comments from the College Quality Team 鈥 Foundation and Cornerstone Standards

Quality of notes and the follow-up and management of test results seem to be high on the radar of the HDC at the moment and is something that is often not prioritised by practices until something goes wrong.

The three issues here seem to be the management of test results, continuity of care and system processes (of which many are outside the scope of a general practice).

During the recent Foundation Standard update the team strengthened the requirements around these areas.

The College recognises the complexity of effectively managing clinical correspondence in general practice and has included a more robust, overarching approach to clinical management in the College鈥檚 quality assurance programme, Foundation Standard 2022.

The standard sets out the requirement for each practice to have documented policies and processes for the transfer of patient clinical responsibility, clinical correspondence, investigation, and clinical governance. Having documented policies and processes is especially important when many members of the practice team have responsibilities in the management of patient test results. Practices can also take some immediate steps to mitigate the risk of errors in their processes by;

  • Making sure all team members are familiar with the practice clinical processes including locums (refresh at a practice meeting).聽
  • Looking at what safety netting measures are in place for GPs inbox management when GPs are away and at GP handovers.
  • The practice clinical governance group could review and update their current policies and processes to align with Foundation Standard 2022 version.
  • Check clinical governance reporting lines are functional and that regular auditing activities are in place for clinical correspondence management.
  • Consider a patient portal allowing patients access to their results as the usual features.