Implementation of ask-advise-connect for smoking cessation in Dutch general practice during the COVID-19 pandemic: a mixed-methods evaluation using the CFIR framework.

BACKGROUND: The Ask-Advise-Connect (AAC) approach can help primary care providers to increase the number of people who attempt to quit smoking and enrol into cessation counselling. We implemented AAC in Dutch general practice during the COVID-19 pandemic. In this study we describe how AAC was received in Dutch general practice and assess which factors played a role in the implementation. METHODS: A mixed-methods approach was used to evaluate the implementation of AAC. Implementation took place between late 2020 and early 2022 among 106 Dutch primary care providers (general practitioners (GPs), practice nurses and doctor's assistants). Quantitative and qualitative data were collected through four online questionnaires. A descriptive analysis was conducted on the quantitative data. The qualitative data (consisting of answers to open-ended questions) were inductively analysed using axial codes. The Consolidated Framework for Implementation Research was used to structure and interpret findings. RESULTS: During the study, most participants felt motivated (84-92%) and able (80-94%) to apply AAC. At the end of the study, most participants reported that the AAC approach is easy to apply (89%) and provides advantages (74%). Routine implementation of the approach was, however, experienced to be difficult. More GPs (30-48%) experienced barriers in the implementation compared to practice nurses and doctor's assistants (7-9%). The qualitative analysis showed that especially external factors, such as a lack of time or priority to discuss smoking due to the COVID-19 pandemic, negatively influenced implementation of AAC. CONCLUSIONS: Although AAC was mostly positively received in Dutch general practice, implementation turned out to be challenging, especially for GPs. Lack of time to discuss smoking was a major barrier in the implementation. Future efforts should focus on providing additional implementation support to GPs, for example with the use of e-health.

Using safety culture results to guide the merger of four general practices in the UK.

BACKGROUND: The rapid merger in a crisis of three GP practices to incorporate the patients from a neighbouring closing surgery, led to the redesign of primary care provision. A deliberate focus on patient safety and staff engagement was maintained throughout this challenging transition to working at scale in an innovative, integrated and collaborative GP model. METHOD: 3 cycles of a staff culture tool (Safety, Communication, Organizational Reliability, Physician & Employee burn-out and Engagement) were performed at intervals of 9-12 months with structured feedback and engagement with staff after each round. The impact of different styles of feedback, the effect of specific interventions, and overall changes in safety climate and culture domains were observed in detail throughout this time period. RESULTS: Strong themes demonstrated were that: there was a general improvement in all culture domains; specific focus on teams that expressed they were struggling created the most effective outcomes; an initial lack of trust of the management structure improved; adapting and tailoring the styles of feedback was most efficacious; and burn-out scores dropped progressively. A unique observation of the rate at which different modalities of safety climate and culture change with time is demonstrated. CONCLUSION: With limited time, resources and energy, especially at times of crisis or change, the rapid and accurate identification of which domains of 'culture' and which teams required the most input at each stage of the journey is invaluable. Using this tool and prioritising patient safety, enables rapid and effective positive change to the culture and shape of expanding practices. It affirms that new models of working at scale in GP can be positively embraced with improvements in safety culture, if this is deliberately focused on and included in the transition process.

Clínica médica general: norma técnica No. 17-2020-DRACES / General medical clinic: technical standard No. 17-2020-DRACES

DRACES [Departamento de Regulación, Acreditación y Control de Establecimientos de Salud] Este documento tiene como objeto: "la Regulación, Autorización y Control de las Clínicas Médicas Generales, en concordancia con el Reglamento para la Regulación, Autorización, Acreditación y Control de Establecimientos de Atención para la Salud, Acuerdo Gubernativo 376-2007." Es de carácter obligatorio, por lo que se aplica en todo el territorio nacional. Contiene además, las definiciones de los conceptos relacionados al tema principal, además de la infraestructura que deberá tener cada clínica, incluidos el equipo y recurso humano y técnico.

Variation in the implementation of PaTz: a method to improve palliative care in general practice - a prospective observational study.

BACKGROUND: PaTz (palliative care at home) is a method to improve palliative care in the primary care setting in the Netherlands. PaTz has three basic principles: (1) local GPs and DNs meet at least six times per year to identify and discuss their patients with a life-threatening illness; (2) these meetings are supervised by a specialist palliative care professional; (3) groups use a palliative care register on which all identified patients are listed. Since the start in 2010, the number of PaTz-groups in the Netherlands has been growing consistently. Although the theory of all PaTz-groups is the same, the practical functioning of PaTz-groups may vary substantially, which may complicate further implementation of PaTz as well as interpretation of effect studies. This study aims to describe the variation in practice of PaTz-groups in the Netherlands. METHOD: In this prospective observational study, ten PaTz-groups logged and described the activities in their meetings as well as the registered and discussed patients and topics of discussions in registration forms for a 1 year follow-up period. In addition, non-participatory observations were performed in all participating groups. Meeting and patient characteristics were analysed using descriptive statistics. Conventional content analysis was performed in the analysis of topic discussions. RESULTS: While the basic principles of PaTz are found in almost every PaTz-group, there is considerable variation in the practice and content of the meetings of different PaTz-groups. Most groups spend little time on other topics than their patients, although the number of patients discussed in a single meeting varies considerably, as well as the time spent on an individual patient. Most registered patients were diagnosed with cancer and patient discussions mainly concerned current affairs and rarely concerned future issues. CONCLUSION: The basic principles are the cornerstone of any PaTz-group. At the same time, the observed variation between PaTz-groups indicates that tailoring a PaTz-group to the needs of its participants is important and may enhance its sustainability. The flexibility of PaTz-groups may also provide opportunity to modify the content and tools used, and improve identification of palliative patients and advance care planning.

[The general practitioner's knowledge and attitude towards proton pump inhibitors adverse effects]. / Connaissances et attitudes des médecins généralistes à l'égard des effets indésirables des inhibiteurs de la pompe à protons.

INTRODUCTION: Proton pump inhibitors (PPIs) are frequently prescribed by general practitioners (GPs) who consider these drugs to be safe. However, more and more adverse drug reactions (ADRs) associated with PPIs are described. OBJECTIVE: To determine the level of knowledge and attitude of GPs with respect to the ADRs of PPIs in adults. METHOD: The GPs of the Nouvelle-Aquitaine region and the chief residents or former chief residents of the 35 French university departments of general medicine were questioned online using a list of 12 ADRs based on a semi-systematic literature review (neutropenia, thrombocytopenia, anemia, hypomagnesemia, increased liver enzymes, renal failure, pneumonitis, skin reactions, fracture, gastric polyps, microscopic colitis, Clostridium difficile colitis) and their attitude towards these ADRs. RESULT: Out of the 232 (7.5%) GPs who participated, 38.4% had come across PPI ADRs (mainly digestive) in the last 6 months. These ADRs had been reported to pharmacovigilance in only 2.3% of the cases. No ADR at all was known by more than 70% of the GPs. The attitude of the GPs who had identified them was, in the majority, to stop PPI treatement. No difference in knowledge was found between university supervisors and non-supervisors. Finally, 74.7% of GPs declared that they required more information on IPP ADRs and 80.4% were willing to "deprescribe" these drugs. CONCLUSION: Aside from digestive ADRs, GPs have a poor knowledge of IPP-related ADRs. Those who do know them have an appropriate attitude and interrupt IPP treatment. The information requested by the GPs about the risks of PPIs is pertinent, particularly in a PPI-adapted prescription procedure, which also involves knowing how to "deprescribe".

A systematic review of patient complaints about general practice.

BACKGROUND: Health care complaints are an underutilized resource for quality and safety improvement. Most research on health care complaints is focused on secondary care. However, there is also a need to consider patient safety in general practice, and complaints could inform quality and safety improvement. OBJECTIVE: This review aimed to synthesize the extant research on complaints in general practice. METHODS: Five electronic databases were searched: Medline, Web of Science, CINAHL, PsycINFO and Academic Search Complete. Peer-reviewed studies describing the content, impact of and motivation for complaints were included and data extracted. Framework synthesis was conducted using the Healthcare Complaints Analysis Tool (HCAT) as an organizing framework. Methodological quality was appraised using the Quality Assessment Tool for Studies with Diverse Designs (QATSDD). RESULTS: The search identified 2960 records, with 21 studies meeting inclusion criteria. Methodological quality was found to be variable. The contents of complaints were classified using the HCAT, with 126 complaints (54%) classified in the Clinical domain, 55 (23%) classified as Management and 54 (23%) classified as Relationships. Motivations identified for making complaints included quality improvement for other patients and monetary compensation. Complaints had both positive and negative impacts on individuals and systems involved. CONCLUSION: This review highlighted the high proportion of clinical complaints in general practice compared to secondary care, patients' motivations for making complaints and the positive and negative impacts that complaints can have on health care systems. Future research focused on the reliable coding of complaints and their use to improve quality and safety in general practice is required.

How older patients prioritise their multiple health problems: a qualitative study.

BACKGROUND: Patients with multimorbidity often receive diverse treatments; they are subjected to polypharmacy and to a high treatment burden. Hence it is advocated that doctors set individual health and treatment priorities with their patients. In order to apply such a concept, doctors will need a good understanding of what causes patients to prioritise some of their problems over others. This qualitative study explores what underlying reasons patients have when they appraise their health problems as more or less important. METHODS: We undertook semi-structured interviews with a purposive sample of 34 patients (aged 70 years and over) in German general practices. Initially, patients received a comprehensive geriatric assessment, on the basis of which they rated the importance of their uncovered health problems. Subsequently, they were interviewed as to why they considered some of their problems important and others not. Transcripts were analysed using qualitative content analysis. RESULTS: Patients considered their health problems important, if they were severe, constant, uncontrolled, risky or if they restricted daily activities, autonomy and social inclusion. Important problems often correlated with negative feelings. Patients considered problems unimportant, if they were related to a bearable degree of suffering, less restrictions in activities, or psychological adjustment to diseases. Altogether different reasons occurred on the subject of preventive health issues. CONCLUSIONS: Patients assess health problems as important if they interfere with what they want from life (life values and goals). Psychological adjustment, by contrast, facilitates a downgrading of the importance. Asking patients with multimorbidity, which health problems are important, may guide physicians to treatment priorities and health problems in need of empowerment.

The effectiveness of repeating a social norm feedback intervention to high prescribers of antibiotics in general practice: a national regression discontinuity design.

OBJECTIVES: Unnecessary antibiotic prescribing contributes to antimicrobial resistance. A randomized controlled trial in 2014-15 showed that a letter from England's Chief Medical Officer (CMO) to high-prescribing GPs, giving feedback about their prescribing relative to the norm, decreased antibiotic prescribing. The CMO sent further feedback letters in succeeding years. We evaluated the effectiveness of the repeated feedback intervention. METHODS: Publicly available databases were used to identify GP practices whose antibiotic prescribing was in the top 20% nationally (the intervention group). In April 2017, GPs in every practice in the intervention group (n=1439) were sent a letter from the CMO. The letter stated that, 'the great majority of practices in England prescribe fewer antibiotics per head than yours'. Practices in the control group received no communication (n=5986). We used a regression discontinuity design to evaluate the intervention because assignment to the intervention condition was exogenous, depending on a 'rating variable'. The outcome measure was the average rate of antibiotic items dispensed from April 2017 to September 2017. RESULTS: The GP practices who received the letter changed their prescribing rates by -3.69% (95% CI=-2.29 to -5.10; P<0.001), representing an estimated 124 952 fewer antibiotic items dispensed. The effect is robust to different specifications of the model. CONCLUSIONS: Social norm feedback from a high-profile messenger continues to be effective when repeated. It can substantially reduce antibiotic prescribing at low cost and on a national scale. Therefore, it is a worthwhile addition to antimicrobial stewardship programmes.

Never events in general practice: a focus group study exploring the views of English and Scottish general practitioners of 'never events'.

OBJECTIVES: To examine general practitioner (GP) understanding of the never event (NE) concept in general practice, and to identify potential enablers and barriers to implementation in UK general practice. DESIGN: Qualitative study using focus groups. The data were analysed thematically and were informed by the normalisation process theory. SETTING: General practice in Northwest England and Southwest Scotland. PARTICIPANTS: 25 GPs took part in five focus groups. 13 GPs were female and 12 male with an age range of 28-60. RESULTS: The NE approach of avoiding serious preventable adverse outcomes from healthcare fitted with participants expectations of the delivery of care but the implementation of strategies to prevent the specific NE was considered complex and variable. The main themes identified participants' understandings and perceived limitations of the NE concept; the embedded layers of responsibility to implement NE within practices and the work required for implementation within general practices. Participants' accounts highlighted the differential nature of work in general practice and that the implementation of initiatives to address specific NE should be situated within a learning and systems approach to implementation. Some NEs were considered more relevant and amenable to simple solutions than others which could influence implementation. CONCLUSIONS: The NE concept was considered overall an important approach to help address key primary care patient safety issues. The utility of individual NEs may vary depending on the complexity of the initiatives that would be needed to manage related risks to as low as reasonably practicable.

[Determinants of time required by medical information technicians for quality control of hospital activity coding, in French medico-administrative system]. / Déterminants du temps nécessaire au contrôle qualité dans le cadre du programme de médicalisation des systèmes d'information (PMSI), dans le champ médecine, chirurgie et obstétrique (MCO) d'un centre hospitalier régional, par les techniciens d'information médicale.

BACKGROUND: Since 2008, in France, hospital funding is determined by the nature of activities provided (activity-based funding). Quality control of hospital activity coding is essential to optimize hospital remuneration. There is a need for reliable tools to allocate human resources wisely in order to improve these controls. METHODS: The main objective of this study was to identify the determinants of time needed by medical information technicians to control hospital activity coding in a Regional Hospital Center. From March 2016 to the beginning of January 2017, medical information technicians reported the time they spent on each quality control, and the time they needed when they had to code the entire stay. Multiple linear regressions were performed to identify the determinants of quality control or coding duration. A split sample validation was used: model was created on one half of the sample and validated on the remaining half. RESULTS: Among the controls, 5431 were included in the analysis of determinants of control duration (2715 kept aside for model validation). Seven determinants have been identified (stay duration, level of complexity, month of control, type of control, medical information technician, rank of classing information, and major diagnostic category). The correlation coefficient between predicted and real control duration was 0.71 (P<10-4); 808 stays were included in the analysis of determinants of coding duration (404 kept aside for model validation). Two determinants have been identified. The correlation coefficient, between predicted and real coding duration, was 0.47 (P<10-3). We performed the same multiple regression, on 2017 activity data, to estimate the weight of each hospital activity pole, regarding quality control of hospital activity coding. CONCLUSION: We succeeded in modeling time needed for quality control of hospital stays. These results helped to estimate human resources required for quality control of each hospital pole. Nevertheless, the second analysis did not give satisfactory results: we failed in modeling time needed to code hospital stays.

"The median isn't the message": How to communicate the uncertainties of survival prognoses to cancer patients in a realistic and hopeful way.

This study investigated how doctors communicate the uncertainties of survival prognoses to patients recently diagnosed with life-threatening cancer, and suggests ways to improve this communication. Two hundred thirty-eight Norwegian oncologists and general practitioners (GPs) participated in Study 1. The study included both a scenario and a survey. The scenario asked participants to respond to a hypothetical patient who wanted to know how long (s)he could be expected to live. There were marked differences in responses within both groups, but few differences between the GPs and oncologists. There was a strong reluctance among doctors to provide patients with a prognosis. Even when they were presented with a statistically well-founded right-skewed survival curve, only a small minority provided hope by communicating the variation in survival time. In Study 2, 177 healthy students rated their preferences for different ways of receiving information regarding the uncertainty of a survival prognosis. Participants who received an explicitly described right-skewed survival curve believed that they would feel more hopeful. These participants also obtained a more realistic understanding of the variation in survival than those who did not receive this information. Based on the findings of the two studies and on extant psychological research, the author suggests much-needed guidelines for communicating survival prognoses in a realistic and optimistic way to patients recently diagnosed with life-threatening cancer. In particular, the guidelines emphasise that the doctor explains the often strongly right-skewed variation in survival time, and thereby providing the patient with realistic hope.

Impact of GP gatekeeping on quality of care, and health outcomes, use, and expenditure: a systematic review.

BACKGROUND: GPs often act as gatekeepers, authorising patients' access to specialty care. Gatekeeping is frequently perceived as lowering health service use and health expenditure. However, there is little evidence suggesting that gatekeeping is more beneficial than direct access in terms of patient- and health-related outcomes. AIM: To establish the impact of GP gatekeeping on quality of care, health use and expenditure, and health outcomes and patient satisfaction. DESIGN AND SETTING: A systematic review. METHOD: The databases MEDLINE, PreMEDLINE, Embase, and the Cochrane Library were searched for relevant articles using a search strategy. Two authors independently screened search results and assessed the quality of studies. RESULTS: Electronic searches identified 4899 studies (after removing duplicates), of which 25 met the inclusion criteria. Gatekeeping was associated with better quality of care and appropriate referral for further hospital visits and investigation. However, one study reported unfavourable outcomes for patients with cancer under gatekeeping, and some concerns were raised about the accuracy of diagnoses made by gatekeepers. Gatekeeping resulted in fewer hospitalisations and use of specialist care, but inevitably was associated with more primary care visits. Patients were less satisfied with gatekeeping than direct-access systems. CONCLUSION: Gatekeeping was associated with lower healthcare use and expenditure, and better quality of care, but with lower patient satisfaction. Survival rate of patients with cancer in gatekeeping schemes was significantly lower than those in direct access, although primary care gatekeeping was not otherwise associated with delayed patient referral. The long-term outcomes of gatekeeping arrangements should be carefully studied before devising new gatekeeping policies.

Efficiency versus thoroughness in medication review: a qualitative interview study in UK primary care.

BACKGROUND: Medication reviews may improve the safety of prescribing and the National Institute for Health and Care Excellence (NICE) highlights the importance of involving patients in this process. AIM: To explore GP and pharmacist perspectives on how medication reviews were conducted in general practice in the UK. DESIGN AND SETTING: Analysis of semi-structured interviews with GPs and pharmacists working in the South West of England, Northern England, and Scotland, sampled for heterogeneity. Interviews took place between January and October 2017. METHOD: Interviews focused on experience of medication review. Data saturation was achieved when no new insights arose from later interviews. Interviews were analysed thematically. RESULTS: In total, 13 GPs and 10 pharmacists were interviewed. GPs and pharmacists perceived medication review as an opportunity to improve prescribing safety. Although interviewees thought patients should be involved in decisions about their medicines, high workload pressures meant that most medication reviews were conducted with limited or no patient input. For some GPs, a medication review was done 'in the quickest way possible to say that it was done'. Pharmacists were perceived by both professions as being more thorough but less time efficient than GPs, and few pharmacists were routinely involved in medication reviews even in practices employing a pharmacist. Interviewees argued that it was easier to continue medicines than it was to stop them, particularly because stopping medicines required involving the patient and this generated extra work. CONCLUSION: Practices tended to prioritise being efficient (getting the work done) rather than being thorough (doing it well), so that most medication reviews were carried out with little or no patient involvement, and medicines were rarely stopped or reduced. Time and resource constraints are an important barrier to implementing NICE guidance.

An Australian community jury to consider case-finding for dementia: Differences between informed community preferences and general practice guidelines.

BACKGROUND: Case-finding for dementia is practised by general practitioners (GPs) in Australia but without an awareness of community preferences. We explored the values and preferences of informed community members around case-finding for dementia in Australian general practice. DESIGN, SETTING AND PARTICIPANTS: A before and after, mixed-methods study in Gold Coast, Australia, with ten community members aged 50-70. INTERVENTION: A 2-day citizen/community jury. Participants were informed by experts about dementia, the potential harms and benefits of case-finding, and ethical considerations. PRIMARY AND SECONDARY OUTCOMES: We asked participants, "Should the health system encourage GPs to practice 'case-finding' of dementia in people older than 50?" Case-finding was defined as a GP initiating testing for dementia when the patient is unaware of symptoms. We also assessed changes in participant comprehension/knowledge, attitudes towards dementia and participants' own intentions to undergo case-finding for dementia if it were suggested. RESULTS: Participants voted unanimously against case-finding for dementia, citing a lack of effective treatments, potential for harm to patients and potential financial incentives. However, they recognized that case-finding was currently practised by Australian GPs and recommended specific changes to the guidelines. Participants increased their comprehension/knowledge of dementia, their attitude towards case-finding became less positive, and their intentions to be tested themselves decreased. CONCLUSION: Once informed, community jury participants did not agree case-finding for dementia should be conducted by GPs. Yet their personal intentions to accept case-finding varied. If case-finding for dementia is recommended in the guidelines, then shared decision making is essential.

Study protocol for a multi-methods study: SAVOIR - evaluation of specialized outpatient palliative care (SAPV) in Germany: outcomes, interactions, regional differences.

BACKGROUND: Since 2007, the German statutory health insurance covers Specialized Outpatient Palliative Care (SAPV). SAPV offers team-based home care for patients with advanced and progressive disease, complex symptoms and life expectancy limited to days, weeks or months. The introduction of SAPV is ruled by a directive (SAPV directive). Within this regulation, SAPV delivery models can and do differ regarding team structures, financing models, cooperation with other care professionals and processes of care. The research project SAVOIR is funded by G-BA's German Innovations Fund to evaluate the implementation of the SAPV directive. METHODS: The processes, content and quality of SAPV will be evaluated from the perspectives of patients, SAPV teams, general practitioners and other care givers and payers. The influence of different contracts, team and network structures and regional and geographic settings on processes and results including patient-reported outcomes will be analyzed in five subprojects: [1] structural characteristics of SAPV and their impact on patient care, [2] quality of care from the perspective of patients, [3] quality of care from the perspective of SAPV teams, hospices, ambulatory nursing services, nursing homes and other care givers, content and extent of care from [4] the perspective of General Practitioners and [5] from the perspective of payers. The evaluation will be based on different types of data: team and organizational structures, treatment data based on routine documentation with electronic medical record systems, prospective assessment of patient-reported outcomes in a sample of SAPV teams, qualitative interviews with other stakeholders like nursing and hospice services, a survey in general practitioners and a retrospective analysis of claims data of all SAPV patients, covered by the health insurance fund BARMER in 2016. DISCUSSION: Data analysis will allow identification of variables, associated with quality of SAPV. Based on these findings, the SAVOIR study group will develop recommendations for the Federal Joint Committee for a revision of the SAPV directive. TRIAL REGISTRATION: German Clinical Trials Register (DRKS): DRKS00013949 (retrospectively registered, 14.03.2018), DRKS00014726 (14.05.2018), DRKS00014730 (30.05.2018). Subproject 3 is an interview study with professional caregivers and therefore not registered in DRKS as a clinical study.

Performance indicators in colonoscopy after certification for independent practice: outcomes and predictors of competence.

BACKGROUND: Robust real-world performance data of newly independent colonoscopists are lacking. In the United Kingdom, provisional colonoscopy certification (PCC) marks the transition from training to newly independent practice. We aimed to assess changes in key performance indicators (KPIs) such as cecal intubation rate (CIR) in the periods pre- and post-PCC, particularly regarding rates and predictors of trainees exhibiting a drop in performance (DIP), defined as CIR <90% in the first 50 procedures post-PCC. METHODS: A prospective United Kingdom-wide observational study of Joint Advisory Group on Gastrointestinal Endoscopy Electronic Training System (JETS) e-portfolio colonoscopy entries (257,800) from trainees awarded PCC between July 2011 and 2016 was undertaken. Moving average analyses were used to study KPI trends relative to PCC. Pre-PCC trainee, trainer, and training environment factors were compared between DIP and non-DIP cohorts to identify predictors of DIP. RESULTS: Seven hundred thirty-three trainees from 180 centers were awarded PCC after a median of 265 procedures and 3.1 years. Throughout the early post-PCC period, average CIRs surpassed the national 90% standard. Despite this, not all trainees achieved this standard post-PCC, with DIP observed in 18.4%. DIP was not influenced by trainer presence and diminished after 100 additional procedures. On multivariable analysis, pre-PCC CIRs and trainer specialty were predictive of DIP. Trainees with DIP incurred higher post-PCC rates of moderate to severe discomfort despite requiring higher analgesic dosages and were more likely to require trainer assistance in failed procedures. CONCLUSIONS: The current PCC requirements are appropriate for diagnostic colonoscopy. It is possible to identify predictors of underperformance in trainees, which may be of value to training leads and could improve the patient experience.

Associations between general practice characteristics with use of urgent referrals for suspected cancer and endoscopies: a cross-sectional ecological study.

BACKGROUND: Large variation in measures of diagnostic activity has been described previously between English general practices, but related predictors remain understudied. OBJECTIVE: To examine associations between general practice population and characteristics, with the use of urgent referrals for suspected cancer, and use of endoscopy. METHODS: Cross-sectional observational study of English general practices. We examined practice-level use (/1000 patients/year) of urgent referrals for suspected cancer, gastroscopy, flexible sigmoidoscopy and colonoscopy. We used mixed-effects Poisson regression to examine associations with the sociodemographic profile of practice populations and other practice attributes, including the average age, sex and country of qualification of practice doctors. RESULTS: The sociodemographic characteristics of registered patients explained much of the between-practice variance in use of urgent referrals (32%) and endoscopic investigations (18-25%), all being higher in practices with older and more socioeconomically deprived patients. Practice-level attributes explained a substantial amount of between-practice variance in urgent referral (19%) but little of the variance in endoscopy (3%-4%). Adjusted urgent referral rates were higher in training practices and those with younger GPs. Practices with mean doctor ages of 41 and 57 years (at the 10th/90th centiles of the national distribution) would have urgent referral rates of 24.1 and 19.1/1000 registered patients, P < 0.001. CONCLUSION: Most between-practice variation in use of urgent referrals and endoscopies seems to reflect health need. Some practice characteristics, such as the mean age of GPs, are associated with appreciable variation in use of urgent referrals, though these associations do not seem strong enough to justify targeted interventions.