Home

Bluesky · LinkedIn

Category: Journal Articles

  • Enhanced Glucose Processing in Gestational Diabetes: Shifting the Diagnostic Landscape

    Jones DL, Kusinski LC, Barker P, Burling K, Halsall I, Turner E, et al. Enhanced glucose processing in gestational diabetes diagnosis: Effects on health equity and clinical outcomes. Diabet Med. 2024 Dec 17;e15476.

    The diagnosis of gestational diabetes mellitus (GDM) has always been a fraught territory. Relying on the oral glucose tolerance test (OGTT), we know the test is inconvenient, variably reproducible, and often poorly tolerated by women. What Jones et al. (2024) add with their OPHELIA study is a careful look at what happens when the pre-analytical handling of blood samples is tightened up: putting samples on ice, rapid centrifugation, and freezing within 2.5 hours. This stops the metabolism of glucose by red cells, leading to. Greater accuracy. The result? Glucose values rise by about 0.6 mmol/L – and the rate of GDM diagnosis leaps from 9% to 22% .

    This is an important finding. The additional women picked up weren’t just at the margins. They were, on average, younger, had higher BMIs, and gave birth to significantly more large-for-gestational-age (LGA) infants (37% vs. 22% under standard testing) . In other words, women with clinically meaningful risk were slipping through the diagnostic net when standard processing was used.

    The equity dimension

    What makes this paper particularly important for those of us in primary care is the way diagnosis intersects with equity. Women most likely to be missed on standard processing were often from higher BMI groups and with differing ethnic profiles. This matters because we already know attendance at OGTT appointments is lower in women from deprived and minoritised ethnic groups . So, the current pathway risks a double inequity: first, some women are less likely to attend, and second, those who do attend may have their hyperglycaemia under-recognised if the sample processing is suboptimal.

    As GPs, we often see the knock-on effects of these diagnostic gaps. Babies born large-for-gestational-age, women with unrecognised hyperglycaemia who may face higher risks of future type 2 diabetes, and families who shoulder the longer-term metabolic consequences.

    But what about overdiagnosis?

    Of course, increasing the diagnosis rate so dramatically raises another concern: are we at risk of overmedicalising pregnancy? Some women identified by enhanced processing required no treatment beyond lifestyle advice, and their risk profile was arguably milder. Jones et al. acknowledge this, but importantly they found the undiagnosed group were not uniformly “low risk” – they had comparable fasting glucose levels to the standard GDM group and higher rates of adverse outcomes.

    For primary care, this tension between under- and over-diagnosis is familiar territory. Just as with hypertension thresholds or cholesterol risk calculators, the line between early intervention and unnecessary burden is a fine one. The challenge will be whether health systems can support the increase in GDM diagnoses with proportionate and person-centred management, without overwhelming services or fuelling unnecessary anxiety in women.

    Implications for practice

    For GPs, the study prompts some key reflections:

    Equity of access: If enhanced processing becomes standard, are we simultaneously addressing the barriers that keep some women from being tested in the first place? Otherwise, the inequity gap may actually widen. Holistic risk framing: Enhanced processing identifies women at higher risk of adverse outcomes – but diagnosis alone isn’t the endpoint. Supporting women through diet, activity, and psychosocial support may be more impactful than simply labelling. The long view: Diagnosing GDM is not just about pregnancy outcomes. It’s also about recognising long-term metabolic risk and ensuring women get follow-up beyond the six-week postnatal check.

    Ultimately, this paper reinforces a core truth: how we handle samples in the lab isn’t a technicality – it shapes who gets a diagnosis, who gets treatment, and who carries risk unnoticed. If we are serious about tackling health inequities in pregnancy, then perhaps the starting point is as simple (and as complex) as a tube of blood and some ice.

  • Understanding the Role of PAPP-A in Pregnancy Outcomes: What the Research Tells Us

    I love PUNS and DENS (Patient Unmet Needs and Doctors Educational Needs). It is something I intend to continue until I stop practising medicine, encourages self-reflection, and is an important part of lifelong learning. I recently had a patient who was having additional screening as they had a low level of PAPP-A discovered at routine antenatal screening. She was consulting about something else, but I wasn’t 100% sure of the role of low PAPP-A levels, and poorer outcomes in pregnancy such as pre-eclampsia and low birth weight. We do very little antenatal care in Defence, it being almost exclusively the preserve of the community midwifery team. I was also unsure of the predictive value of the test so, being inquisitive, did a quick literature review. There were a number of papers I found, but this systematic review and meta-analysis seemed high quality and originated from my affiliated University, Birmingham:

    https://research.birmingham.ac.uk/en/publications/association-of-serum-papp-a-levels-in-first-trimester-with-small-

    I have read it but for the purposes of this blog, asked ChatGPT to summarise the discussion:

    Recent research has explored how a specific protein in early pregnancy, called Pregnancy-Associated Plasma Protein A (PAPP-A), might help predict certain pregnancy complications. PAPP-A levels are measured during the first trimester, and this study found that very low levels of this protein can be linked to a higher risk of complications like preterm birth or having a small baby.

    However, it’s important to note that while low PAPP-A levels do indicate a higher risk, the prediction isn’t perfect. Most women with low levels of PAPP-A will still go on to have healthy pregnancies, and many who experience complications will have normal levels of this protein.

    The study’s strength lies in its robust and thorough approach to reviewing existing research, but there were some limitations. For example, differences in the studies they analyzed led to some inconsistent results. Additionally, the researchers had to simplify their analysis by treating PAPP-A levels as either “low” or “not low,” which can oversimplify the true relationship between PAPP-A and pregnancy outcomes.

    The research highlights that while PAPP-A can be a useful marker, it’s not enough on its own to predict pregnancy outcomes. More research, especially studies that look at individual patients’ data, is needed to fully understand how PAPP-A interacts with other factors throughout pregnancy.

    Currently, in the UK, PAPP-A is used as part of screening for Down syndrome, but it’s not yet used to predict other pregnancy risks. Before it could be used in this way, doctors would need to be sure that they can effectively intervene to help women identified as high-risk based on PAPP-A levels. For now, the study helps doctors provide better information and support to expectant mothers who have low PAPP-A levels, ensuring they understand what it might mean for their pregnancy.

    So there you have it. As I thought, it’s one of those tests where both the sensitivity and specificity are not amazing, but not bad. Indeed, the relationship between placental thickness and PAPP-A levels demonstrated a sensitivity of around 70% and a specificity of around 50% in one study I found. Another study looked at low PAPP-A levels and risk of pre-eclampsia, finding a significant association with low levels (<10th centile). It should be noted, the analysis in this study was not multivariable so there was a risk of confounding variables also having an effect, a point the authors recognise.

    I found this patient information leaflet which summarises it nicely.

  • Improvements in Stress, Affect, and Irritability Following Brief Use of a Mindfulness-based Smartphone App: A Randomized Controlled Trial – Economides et al, 2018

    Available here – full text.

    One area which I reviewed as part of my foundation module in pain was Mindfulness and Mindfulness-Based Stress Reduction (MBSR) to see if there was any benefit for chronic pain sufferers. From what I found in a number of papers I reviewed was that pain scores do not change but acceptance of chronic pain improves.  Since reading about this, it is something I have mentioned to my patients as an option, or at least to investigate it.  I always add the caveat ‘it is not for everyone’ but a surprising number of the young soldiers I see with pain (usually musculoskeletal or urological) engage with it.  I reviewed one paper which looked specifically at Headspace (as one of the smartphone app market leaders).  This was an RCT of Head Space, where patients were randomised to either the app or active control, to see if Head Space was superior (Economides et al., 2018).  My non-expert critical ‘take’ on the paper is summarised below. This study used validated outcome measures which demonstrated statistically improved scores in measures of irritability & anxiety.

    Aim / Type of study An RCT comparing Mindfulness to an active control.

     
    Methods A 3rd party recruiting agency (https://www.findparticipants.com) was used to recruit participants. They did NOT have chronic pain and were naïve to both meditation, mindfulness and the Headspace app. They had no psychological illness.  They had to have access to a smartphone. Using a statistical approach, 52 participants were required to achieve significance at the 0.05% level.

    The Headspace meditation mindfulness app was compared against a control group using Headspace audiobook. The intervention group (N=41) was exposed to the first 10 sessions of 10-minute sessions delivered by the Headspace app.  The control (N=28) was 10 sessions of 10 minutes of excerpts from a Mindfulness & Meditation audiobook.  To closely match the intervention, the audiobook was narrated by the same author in a similar manner.  Thus, the only difference between the two groups was the content.

    There were 4 outcome measures:  Stress Overload Scale (SOS) – personal vulnerability and event load, Scale of Positive and Negative Experience (SPANE) and Brief Irritability Test (BITe).  These were assessed by all participants in both groups and paired T-tests were undertaken between them. Dropouts were included in an intention to treat analysis.

    The outcome measures were all reductions in the 4 measures from baseline.  The reduction in scores in the two groups were assessed by an Analysis of Variance test (ANOVA) as well as post hoc t-tests to further characterise differences.  Intention to treat analysis was also included.
    Results / Data In two of the outcome measures (SOS event load and SPANE) there was significant difference between the groups favouring Head Space (P<=0.01).  SOS personal vulnerability was also different between the groups again favouring Head Space (P=<=0.05). There was no difference in the SOS Personal Vulnerability score between the two groups (P=0.09).
    Comments Whilst the size of the groups was different (41 in Headspace vs 28 in Audiobook), they were broadly similar demographically.  There were more females in the Headspace group and the percentage who agreed or strongly agreed meditation could be beneficial was higher (70.8% vs 57.2%), although the authors comment that demographics were not significantly different (x2 test >0.05).  After randomisation and allocation but before baseline assessment, a much higher number of audiobook participants dropped out (48 vs 33) resulting in a smaller group.

    The outcome measures were scores on validated questionnaires with good internal consistency.  All the tests have good test re-test reliability bar the BITe which has not been determined yet.  The participants were randomised by a rigorous approach and blinded to the intervention they were receiving but it is difficult to ascertain if the reviewers were blinded.  The study was all done on-line and were e-mailed the voucher code which restricted their access to the intervention or control content.  By assumption, this is a single-blinded study as the researcher would have known who was receiving the code.

    In terms of the strengths and weaknesses of the study, two of the researchers were involved in the development of the Head Space app.  The possibility of funding bias is raised by this given a potential conflict of interest with a commercial product.  There are three other areas for critical review in the context of the population researched:

    Area of Reflection Implication for chronic pain patients
    The patients involved did not have chronic pain. The findings of this study may not be generalisable.
    The sample may not represent the general population.  In both groups, those who agreed or strongly agreed meditation can be beneficial is over 50% (29/41 and 16/28 for Head Space and control respectively).  This compares to a large survey which found a lifetime prevalence of 5.2% of the use of meditation for health reasons (Cramer et al., 2016). Patients who have previously done well with psychoeducational pain courses (ie an acceptance of MBSR) may well do better with Mindfulness.
    The demographics of the sample cohort are predominantly white/Caucasian and college/university education.  All three of these reduce the applicability of this study to the wider population, a valid point raised by the researchers when assessing the strengths/weakness of their work. Again the generalisability may be challenged.  The Army has a mixed ethnic demography but is still predominantly Caucasian.  Future work I am considering is to see if commonwealth soldiers with chronic pain can get similar benefits.

    To summarise, MBSR can improve affect and reduce depression scores but may not significantly reduce pain scores. It can be delivered successfully via a smartphone app. It is a low-risk intervention which may have some positive benefits.  This paper is a small snapshot with some methodological weaknesses, but it otherwise seems well constructed and I liked the active control.  To the uninitiated who had not used MBSR before, it seems like a good comparator.  It has certainly prompted some discussion amongst our primary care team about the role of apps in the rehabilitation of our injured soldiers.  We hope to look at two or three of these (including Headspace) in this context and get both objective results as well as the qualitative opinion of what the soldier actually thought.

    CRAMER, H., HALL, H., LEACH, M., FRAWLEY, J., ZHANG, Y., LEUNG, B., ADAMS, J. & LAUCHE, R. 2016. Prevalence, patterns, and predictors of meditation use among US adults: a nationally representative survey. Scientific reports,6,36760.

    ECONOMIDES, M., MARTMAN, J., BELL, M. J. & SANDERSON, B. 2018. Improvements in Stress, Affect, and Irritability Following Brief Use of a Mindfulness-based Smartphone App: A Randomized Controlled Trial.Mindfulness,1-10.

     

  • PBSGL within an interprofessional​ group of doctors.​

    A belated post about my article for the Journal of the Royal Army Medical Corps.  It is available here but alas not in full text.  Indeed I am not sure I am allowed to link to a full-text version as that would make it open access and that in turn makes it expensive to publish.

    Anyhow, it was published, and I am clearly delighted that at the ripe old age of 50, I have discovered chapter 3 (or 4) of my career.  I found the whole process of designing the questionnaire, setting it up on a commercial survey website, seeing the responses come in, importing into Nvivo and doing the thematic analysis fascinating.  After this, relating it to what I had read and understood from the literature, what was replicated, what was new.  I was enthralled, and I can see what makes research so exciting (although this wasn’t research, it was a service evaluation).

    I hope I can do some more work on PBSGL and its impact.  Indeed the next area is to incorporate our Practice Nurse cadre into it, so we have a genuinely interprofessional demographic.

    I think the next areas I am going to divert my attention to are how a GP Registrar and/or a GP answers a clinical question (in response to a DEN) and an exploration of the impact of NFCI on a soldier.  Two completely separate things but also aligning with my day job.  It means more literature searches and more protocol writing, but I think both areas will reveal some rich data and stimulate debate.  My hypotheses (if one can have one when undertaking a ‘service evaluation’) are that the former will throw up some concerns and under-utilisation of Web 2.0/SMS in answering clinical questions.  I think the latter will demonstrate a significant psychological overlay in neuropathy (NFCI) with a subsequent effect on ADLs, and that the first line of treatment should be a set number of early CBT/psychological interventions to mitigate against longer-term disability.

  • Hall, Pippa, and Lynda Weaver. “Interdisciplinary education and teamwork: a long and winding road.” Medical education 35.9 (2001): 867-875

    This article looks at education in an interdiscplinary team.  Most people have heard of a multi-disciplinary team where a team of specialists (in their own right) discuss cases.  Think of the MDT approach to hospital based disciplines and particualr complex cancer care.  At the other end of the spectrum is transdisciplinary where there is role blurring within the team members.  This is increasingly the case in primary care with practice nurses taking on more chronic disease management and the use of physician assistants.  Interdisciplinary is probably a more traditional primary care model however where there are distinct roles yet the team works together with a common focus being the patient.  That is not to say this isn’t the case with a MDT but there is more role blurring with interdiscliplinary teams as well as some professional overlap.  It’s what often happens when we discuss patients in primary care as a team looking at all outcomes for a patient.  These could be physical, social or psychological.  The question this article attempts to answer is how do we address the educational needs of that team?

    My professional project is a service evaluation of PBSGL.  This method of CME has been evluated before with generally favourable findings.  My ‘unique selling point’ so to speak is how having a mixed skillset group can enable peer support.  For example, having GPSTs, First5 GPs as well as more experienced GPs.  Whilst this doesn’t fit in with the definitons above, there are some transferable findings.  The article talks about System Issues and Content Issues.  Both are relevant.  Three areas discussed I think are important.

    1. Non-traditional teaching methods such as Problem Based Learning helps focus the team on ‘idea dominance’. This puts teamwork at the fore, with the patient as the focus. This is an important aspect of PBSGL.
    2. Faculty development – who does the education within the team & how do they as educators develop? This role is extremely important in the early days of initiating a programme and it would be easy to become disillusioned if there is apathy towards interdisciplinary education.
    3. Leadership skills – a hugely important yet often overlooked. We probably are demonstrating good leadership in primary care already but may be need to qualify it to allow development. This article pre-dates it but the NHS Academy Medical Leadership Competency Framework is a good start with a self assessment toolkit for those responsible for faculty development and who are facilitators.

    Whilst an older article, it does raise important aspects of how we assure and develop education within the Primary Care Team.  This is perhaps more of an imperative due to the changing roles of the different Primary Care Team members since the article was written and further ‘role blurring’.

  • OSELTAMIVIR TREATMENT FOR INFLUENZA IN ADULTS: A META-ANALYSIS OF RANDOMISED CONTROLLED TRIALS (Dobson et al., 2015)

    I was interested to read about a new drug NICE have approved for HUS which by all accounts is massively expensive.  All sorts of harrumphing followed in the press as well as the medical press about funding of expensive drugs and QALYs.  Personally I think NICE have an important yet unforgiving job to do.  I also do get a little cynical at yet another expensive breast cancer drug coming out which, accompanied by a powerful patient interest lobby, is guaranteed to net the drug company a good return.  I remember reading the first papers on the initial Herceptin trials which were sponsored by big Pharma & thinking ‘this actually isn’t that great’.  The study sample was small, the mortality pretty unchanged though admittedly the disease free interval was extended.  I suppose if you have any cancer, these are small positives.  My mother is currently in the terminal stages of bowel cancer and I know how awful it can be.

    The other side is a drug which may benefit millions of us in the event of a pandemic.  This brings us to Tamiflu, or oseltamivir.  Not so long ago, a meta-analysis published in the BMJ reported that neuraminidase inhibitors at best had a modest effect (Jefferson et al., 2009).  This was something I had always suspected.  The patients I treated seemed to have more side effects than benefit in my admittedly small sample.  I assumed powerful lobbying by Pharma alongside failure to disclose unpublished trial data had won the day.

    It was the other day therefore whilst reading the Torygraph that this article piqued my interest.  I won’t go into massive detail as one needs access to full text to assess it.  Suffice to say it seems a well constructed systematic review with lots of blobograms (Forest plots) and actually is quite easy to read.  The statistics seem sound too.  When I see something that is too good to be true, it usually is.  Not wanting to be unfair on the authors, there are one or two concerns.  Firstly 2 of the 4 authors have been paid by Roche.  Secondly, whilst they had no part in the analysis, Roche funded the study under the guise of the MUGAS foundation.  The authors acknowledge this and make valid comment.  The interesting thing about this study is that it used unpublished data.  I don’t know which studies were unpublished but I could probably have a good guess.  Of the 9 studies included, a sizeable minority have confidence intervals that cross the line of no effect.  This includes the 2nd largest.  Clearly when doing a systematic review, one includes all the studies but it is interesting to see all the data.

    Probably the biggest issue and the one that that relates to primary care I the UK is the different analyses that were carried out.  The intention-to-treat-infected population showed far greater effectiveness than the intention-to-treat population.  This is essentially saying that when the nasal swabs were analysed & the results broken down into those who had influenza & those who didn’t, the former group gained far more benefit from oseltamivir.  Patients included in the studies presented with symptoms of influenza had their treatment started within 36hr of onset. In other words no time to obtain a lab result to confirm influenza.  This means the only population relevant to us in primary care is the intention-to-treat population where the effect is much less.  Indeed one of the headlines is reduction of  hospital admissions which in the intention-to-treat group was not significant.  What we really need is a rapid near test to enable a more accurate diagnosis so the drug is used more effectively.

    There are several references to Jefferson’s earlier work and indeed for anyone interested, I would read both studies together.  It is a good example of a systematic review and actually one of the easiest to read I have come across in a while.  It also will give some justification to the Government for the millions of pounds spent on stockpiling drugs in the event of an influenza pandemic.

    DOBSON, J., WHITLEY, R. J., POCOCK, S. & MONTO, A. S. 2015. Oseltamivir treatment for influenza in adults: a meta-analysis of randomised controlled trials. The Lancet.

    JEFFERSON, T., JONES, M., DOSHI, P. & DEL MAR, C. 2009. Neuraminidase inhibitors for preventing and treating influenza in healthy adults: systematic review and meta-analysis. Bmj, 339.

  • OSELTAMIVIR TREATMENT FOR INFLUENZA IN ADULTS: A META-ANALYSIS OF RANDOMISED CONTROLLED TRIALS (Dobson et al., 2015)

    I was interested to read about a new drug NICE have approved for HUS which by all accounts is massively expensive. All sorts of harrumphing followed in the press as well as the medical press about funding of expensive drugs and QALYs. Personally I think NICE have an important yet unforgiving job to do. I also do get a little cynical at yet another expensive breast cancer drug coming out which, accompanied by a powerful patient interest lobby, is guaranteed to net the drug company a good return. I remember reading the first papers on the initial Herceptin trials which were sponsored by big Pharma & thinking ‘this actually isn’t that great’. The study sample was small, the mortality pretty unchanged though admittedly the disease free interval was extended. I suppose if you have any cancer, these are small positives. My mother is currently in the terminal stages of bowel cancer and I know how awful it can be.

    The other side is a drug which may benefit millions of us in the event of a pandemic. This brings us to Tamiflu, or oseltamivir. Not so long ago, a meta-analysis published in the BMJ reported that neuraminidase inhibitors at best had a modest effect (Jefferson et al., 2009). This was something I had always suspected. The patients I treated seemed to have more side effects than benefit in my admittedly small sample. I assumed powerful lobbying by Pharma alongside failure to disclose unpublished trial data had won the day.

    It was the other day therefore whilst reading the Torygraph that this article piqued my interest. I won’t go into massive detail as one needs access to full text to assess it. Suffice to say it seems a well constructed systematic review with lots of blobograms (Forest plots) and actually is quite easy to read. The statistics seem sound too. When I see something that is too good to be true, it usually is. Not wanting to be unfair on the authors, there are one or two concerns. Firstly 2 of the 4 authors have been paid by Roche. Secondly, whilst they had no part in the analysis, Roche funded the study under the guise of the MUGAS foundation. The authors acknowledge this and make valid comment. The interesting thing about this study is that it used unpublished data. I don’t know which studies were unpublished but I could probably have a good guess. Of the 9 studies included, a sizeable minority have confidence intervals that cross the line of no effect. This includes the 2nd largest. Clearly when doing a systematic review, one includes all the studies but it is interesting to see all the data.

    Probably the biggest issue and the one that that relates to primary care in the UK is the different analyses that were carried out. The intention-to-treat-infected population showed far greater effectiveness than the intention-to-treat population. This is essentially saying that when the nasal swabs were analysed & the results broken down into those who had influenza & those who didn’t, the former group gained far more benefit from oseltamivir. Patients included in the studies presented with symptoms of influenza had their treatment started within 36hr of onset. In other words no time to obtain a lab result to confirm influenza. This means the only population relevant to us in primary care is the intention-to-treat population where the effect is much less. Indeed one of the headlines is reduction of hospital admissions which in the intention-to-treat group was not significant. What we really need is a rapid near test to enable a more accurate diagnosis so the drug is used more effectively.

    There are several references to Jefferson’s earlier work and indeed for anyone interested, I would read both studies together. It is a good example of a systematic review and actually one of the easiest to read I have come across in a while. It also will give some justification to the Government for the millions of pounds spent on stockpiling drugs in the event of an influenza pandemic.

    DOBSON, J., WHITLEY, R. J., POCOCK, S. & MONTO, A. S. 2015. Oseltamivir treatment for influenza in adults: a meta-analysis of randomised controlled trials. The Lancet.

    JEFFERSON, T., JONES, M., DOSHI, P. & DEL MAR, C. 2009. Neuraminidase inhibitors for preventing and treating influenza in healthy adults: systematic review and meta-analysis. Bmj, 339.