Microbiology Dissertation Topics: Narrow, Researchable Ideas for UK Students

Microbiology dissertation topics only work when they name an organism or community, a setting, a measurement and a route to the evidence. Five questions help you test any idea against that standard before you commit.

• Can I get the organism, containment level and bench time this project needs, or should it run on existing data?
• Is my question about one organism or community, in one setting, with one measurable outcome?
• Where will my numbers come from: my own isolates, NHS and UKHSA surveillance tables, or public sequences?
• If resistance is my outcome, which susceptibility standard and which breakpoint version will I cite?
• Will my risk assessment and ethics route be settled early enough to leave time for repeat experiments?

Microbiology is a wide field, so coverage needs a plain statement. Most topics below centre on bacteria and fungi, with some work on phages, viruses and community sequencing. Host-response questions fit better with immunology dissertation topics, and broader life-science ideas sit with biology dissertation topics.

Position correct as of October 2026 for the policy and surveillance items named below: the UK antimicrobial resistance action plan, Candidozyma auris notification, MHRA phage guidance and gonorrhoea resistance reporting. Check UKHSA and the MHRA for the current status before you cite any of them.

Microbiology dissertation topics grouped by evidence route

Topics are grouped by where the evidence comes from, because that decides whether a project finishes. Desk-based work needs no laboratory. Isolate work, however, depends on supervision, containment and reagents. Numbering runs continuously, so each idea can be named in a supervisor meeting.

Desk-based microbiology research topics from UKHSA and NHS data

These need no laboratory, yet they still reward careful method. The annual ESPAUR reports and UKHSA’s bacteraemia commentary are the usual starting sources.

1. Which acute trusts in England fall outside expected variation for E. coli bacteraemia once onset category and population size are taken into account?
UKHSA’s annual epidemiological commentary reports counts and rates by acute trust and sub-integrated care board, and the E. coli series starts in April 2012. A funnel plot or a negative binomial model can show which trusts sit outside expected variation. The main pitfall is reading community-onset infections as a measure of hospital performance, so separate the onset categories first.

2. Has the national MRSA bacteraemia series flattened since April 2007, and where does a segmented regression place the breakpoints?
The commentary carries the MRSA series back to April 2007. A plain undergraduate version plots the series, marks dated national policies and describes change before any modelling. A Master’s version fits an interrupted time series to quarterly counts. Either way, treat policies as hypotheses for a change, never as proven causes.

3. Do community-onset and healthcare-onset Clostridioides difficile infections follow the same trend in England, and what would a divergence imply for prevention?
The same commentary reports C. difficile infection alongside the bacteraemias. Check which onset categories the tables include before designing the comparison. A divergence would point towards different transmission routes, so the discussion needs infection prevention and antibiotic stewardship literature as well as the counts.

4. To what extent do deprivation gradients in antibiotic-resistant bloodstream infections reported in ESPAUR hold for individual organisms?
ESPAUR has reported widening gaps in antimicrobial use and resistant infections between the most and least deprived populations. A project could test whether the gradient is consistent for E. coli, Klebsiella species and Staphylococcus aureus separately, then discuss mechanisms such as exposure, healthcare contact and ascertainment. Because the data are ecological, the aims must not claim individual-level risk.

5. Is England on course for the antimicrobial use targets in the 2024 to 2029 national action plan?
The plan, Confronting antimicrobial resistance 2024 to 2029, was published on 8 May 2024 with indicative targets for March 2029. NHS England reports consumption in defined daily doses per 1,000 inhabitants per day against a 2019 to 2020 baseline. A trajectory analysis compares recent performance with the straight-line path to each target. Targets and milestones may be revised, so cite the current NHS England guidance. Suits undergraduate or Master’s level.

6. How much of the recorded rise in Candidozyma auris cases in England reflects spread, and how much reflects screening and mandatory notification?
The yeast formerly named Candida auris became notifiable on 6 April 2025, and UKHSA publishes six-monthly commentaries by region and specimen type. Screening guidance for acute care also changed in 2025, so headline counts cannot be read as incidence. A strong project separates infections from colonisations, compares regions with different screening intensity and states the ascertainment limits plainly. Search under both names.

7. What does the shift from culture to nucleic acid amplification testing mean for resistance surveillance in Neisseria gonorrhoeae?
The GRASP programme reports resistance in England and Wales from cultured isolates, yet NAAT diagnosis produces no isolate. UKHSA has asked clinics to culture isolates, perform test-of-cure and refer suspected treatment failures. A desk-based project could estimate how many diagnoses can be matched to an isolate and what molecular resistance markers might add. In March 2025 UKHSA reported 42 ceftriaxone-resistant cases since 2015, mostly associated with travel to or from the Asia-Pacific region, so treat later counts as provisional.

8. Do post-pandemic changes in laboratory reports of Campylobacter and Salmonella reflect incidence, or the spread of culture-independent testing?
UKHSA’s report on gastrointestinal infections in England for 2022 to 2024 gives laboratory reports by pathogen. When laboratories adopt multiplex PCR panels, positive reports can rise with no change in illness. Compare pathogens that PCR detects well with those it detects poorly, and read laboratory-method documentation alongside the counts. A Master’s-level design.

9. How much of the fall in vaccine-type invasive pneumococcal disease after PCV13 replaced PCV7 in 2010 was offset by non-vaccine serotypes?
UKHSA surveillance reports serotype-level invasive disease. A descriptive project can compare serotype groups by age band across the years before and after the switch. Use rates rather than counts, because testing practice and population size both changed over the period. This suits undergraduates who want a clear, defensible dataset.

Laboratory microbiology dissertation topics: isolates and diagnostics

These suit students whose supervisor can supply isolates, reference strains or a diagnostic laboratory placement. Check the hazard group of every organism before planning. Diagnostic study designs also appear in pathology dissertation topics.

10. Why does disc diffusion fail for colistin, and how large are the very major errors against broth microdilution in a named Enterobacterales panel?
EUCAST recommends broth microdilution for colistin because disc diffusion is unreliable for the drug. A panel of reference and archived isolates read by both methods lets you report categorical agreement and error rates using the ISO 20776-2 definitions. Include resistant controls if your department holds them. Otherwise, keep claims at phenotype level.

11. Do mecA-only PCR assays and some PBP2a latex tests miss mecC MRSA, and what should follow a cefoxitin-resistant, mecA-negative result?
mecC-positive isolates are cefoxitin resistant, yet an assay built only around mecA can return a negative result. Because mecC is rare, prevalence claims need an archived collection or a reference panel, never a handful of fresh isolates. Frame the work as an assay evaluation with a defined algorithm. PhD chapter scale if it extends to sequencing.

12. Does direct MALDI-TOF identification from positive blood culture bottles shorten time to identification compared with overnight subculture, and at what cost in accuracy?
Design this as a service evaluation using residual material. That route often differs from patient research, so confirm it with the laboratory manager and your ethics committee before collecting anything. Report time to identification, species-level agreement and the share of polymicrobial or low-score results.

13. How many Enterobacterales isolates change category when the same zone diameters are interpreted with EUCAST rather than CLSI breakpoints, and for which drugs?
Many UK laboratories follow EUCAST, after BSAC aligned its method with it, whereas much of the international literature uses CLSI. Students therefore often compare their results with studies that used a different standard. Re-interpreting an archived set of zone diameters shows how far resistance prevalence depends on the standard alone. State the version year of each, since breakpoints are revised annually, and note that EUCAST’s “susceptible, increased exposure” category complicates direct comparison.

14. Which automated urine screening thresholds reduce unnecessary cultures without missing true infections in older adults?
Flow cytometry-based urine analysers give bacterial and white-cell counts that laboratories use to decide whether to culture. A retrospective comparison of analyser results against culture at several thresholds gives sensitivity, specificity and the proportion of cultures avoided. Use confidence intervals, and define the reference standard carefully because mixed growth is often reported as contamination rather than infection.

Hospital surfaces, plumbing and biofilms

15. Which Gram-negative organisms persist in intensive care sink drains, and do they match organisms later found in patients?
Hospital wastewater plumbing has been investigated as a reservoir for carbapenemase-producing Enterobacterales. A feasible design swabs drains in one named unit with permission from infection control and estates, cultures on selective media and compares species and resistance genes with the unit’s clinical isolates. Species-level matching proves very little, so use whole-genome sequencing if it is available. Infection prevention angles also run through infectious disease nursing dissertation topics.

16. Does chlorhexidine tolerance differ between planktonic and biofilm Staphylococcus epidermidis from catheter and wound sources, measured as minimum biofilm eradication concentration?
The Calgary biofilm device and crystal-violet assays answer different questions: viability on one hand, biomass on the other. Choose the one that matches your aim and include sterile and solvent controls. Replicates must be biological, meaning separate overnight cultures, because technical replicates from one culture do not estimate real variation.

17. How strong is the evidence that common surface disinfectants reduce Candidozyma auris contamination in healthcare settings?
A systematic review needs no live organism, which makes it practical for undergraduates and Master’s students alike. The World Health Organization placed the species in its critical priority group of fungal pathogens in 2022. Search carrier tests and in-situ studies, record test standards and contact times, and appraise bias with a reporting framework such as PRISMA. The discussion should separate in vitro efficacy from any effect on transmission.

Water, soil and wastewater

Environmental projects usually fail at sampling design. State where, when and how many samples you will take, and which regulator’s data you will compare against. Wider ideas are collected in environmental science dissertation topics.

18. Do storm overflow spill durations recorded by Event Duration Monitors explain short-term E. coli exceedances at one named English bathing water?
The Environment Agency tests designated bathing waters for E. coli and intestinal enterococci through the bathing season, and classifications draw on four seasons of samples. Water companies publish overflow event duration data. Align each sample date with spill hours in the preceding 24 to 72 hours, then control for rainfall. Treat the lag as an assumption to test, and check which regulations applied in each season you analyse. Suits undergraduate or Master’s level.

19. What abundance of sul1 and intI1 can be measured upstream and downstream of a named wastewater treatment works, and does it differ by season?
Quantitative PCR for sul1 and the class 1 integron gene intI1 is widely used as a proxy for resistance pollution. Include extraction blanks and an inhibition check, normalise to 16S rRNA gene copies and sample on matched days. These markers show potential for gene transfer, not infection risk, and the discussion should say so.

20. Can wastewater metagenomics track carbapenemase genes reliably enough to inform hospital infection control?
A critical review at Master’s level could compare published studies on sensitivity, sequencing depth and sewer transit effects. Low prevalence, DNA that may not come from viable organisms and variable sampling all limit inference. The strongest reviews state which decision each study design could actually support.

21. Does ISO 11731 culture miss Legionella that qPCR detects in building water samples, and what would a duty holder do with the difference?
Legionella pneumophila is a hazard group 2 agent in the HSE Approved List, so approvals and containment must be in place before sampling. Culture is slow and misses cells that will not grow, whereas qPCR detects DNA and cannot separate live from dead cells. A paired-sample design reports agreement and then discusses what each result means under HSE guidance on controlling legionella.

22. How do mcrA and pmoA gene abundances change along a restoration-age gradient in a named upland peatland?
Functional gene qPCR links microbial capacity to methane cycling. Choose sites with known drainage or restoration history, record water table and temperature, and sample several independent patches per age class to avoid pseudoreplication. Fieldwork needs landowner and designation permissions, so allow lead time. A Master’s or PhD-scale project.

23. Which methanogen lineages are associated with stable biogas yield when food waste feedstock changes, using published digester time series?
Public repositories hold 16S rRNA gene and metagenomic data from anaerobic digesters. Reanalysing them with one pipeline, plus a record of feedstock and operating parameters, avoids reactor costs. Different primers and platforms create batch effects, so model study as a random effect. For wider waste ideas, see waste management dissertation topics.

24. At what ciprofloxacin concentration does an effluent-relevant dose begin to select resistant E. coli in competition assays?
Minimal selective concentration studies compare the growth of resistant and susceptible strains across concentrations below the MIC. Use hazard group 1 strains with a marked resistance, and report the lowest concentration with a detectable fitness difference. Relate it to measured environmental concentrations from cited studies instead of assuming a typical value.

Food, farm and field

25. What is the growth potential of Listeria monocytogenes in one named chilled product at realistic domestic fridge temperatures?
ISO 20976-1 describes how to design challenge tests for growth potential. Use a defined inoculum, several storage temperatures including some above 5 °C, and a product with known pH and water activity. Because the organism is hazard group 2, a food microbiology laboratory with approval is essential. Suits Master’s level.

26. Does rinsing raw chicken spread bacteria beyond the sink, tested with a non-pathogenic surrogate?
The Food Standards Agency advises against washing raw poultry. A marked E. coli K-12 strain applied to the surface lets you measure splash deposition on nearby surfaces without handling Campylobacter. Justify why the surrogate behaves like the target, and state that results cannot be read as Campylobacter counts. A practical project for undergraduates.

27. How closely do UK-VARSS antibiotic sales trends track resistance in E. coli from food-producing animals?
UK-VARSS, from the Veterinary Medicines Directorate, reports veterinary antibiotic sales and resistance surveillance. Aligning sales by species with resistance proportions over time is feasible as a desk-based project. Sales are not use, and sampling design changes between years, so both limits belong in the aims.

28. How does the ratio of lactic acid bacteria to yeast change over the first fourteen days of a sourdough starter at two hydration levels?
In plain terms, you grow a starter, count its two main microbe groups on selective plates and track acidity. Keep flour, room temperature and feeding times constant, and record them. Use a repeated-measures analysis rather than comparing only the first and last day. An accessible undergraduate laboratory project with hazard group 1 organisms.

29. What proportion of questing Ixodes ricinus nymphs at a named woodland carry Borrelia burgdorferi sensu lato by qPCR?
Check what UKHSA tick surveillance has already published for your region before designing sampling. Blanket dragging needs landowner permission and repeated visits. With few ticks, confidence intervals will be wide, so plan how many survey days you can realistically cover. Borrelia work also needs a containment check in the Approved List.

Fungi and yeasts

30. How common are azole-resistant Aspergillus fumigatus isolates carrying TR34/L98H in compost and plant-waste sites?
Environmental azole use is one proposed route to resistance. Culture on azole-containing agar, then confirm genotype by cyp51A sequencing. The species is hazard group 2, and compost handling carries spore exposure risk. A realistic scope is one or two sites with several sampling dates.

31. Do archived Candida parapsilosis bloodstream isolates carry ERG11 mutations that explain fluconazole resistance?
Pair broth microdilution against the EUCAST method with ERG11 sequencing. The pitfall is a collection too small to contain resistant isolates, so confirm what the laboratory holds before committing. Suits Master’s level.

Phages and alternatives to antibiotics

32. How would a personalised phage therapy pathway work under the MHRA’s 2025 guidance, compared with the pathway for a pre-made phage product?
The MHRA published its first phage guidance on 4 June 2025. It notes that no phage medicines are licensed in the UK and that some patients received imported phages under compassionate use. A desk-based regulatory comparison maps evidence, manufacturing and supply requirements for each pathway. Draft EMA guidance on phage medicine quality was open for consultation in October 2025, so read the MHRA’s current text before writing. Suits Master’s or PhD level.

33. Which phages can be isolated from sewage influent against a named hazard group 1 or 2 host strain, and how broad is their host range across a defined panel?
UKHSA is developing a phage collection, so check whether deposited phages already cover your host. Enrichment, plaque purification and spot tests on a panel give host range. Screen for temperate phages, because they make poor therapy candidates. Sewage handling needs its own risk assessment.

34. Does sub-inhibitory ciprofloxacin enhance phage killing of a defined E. coli strain, and does it also select resistance faster?
Phage-antibiotic synergy appears in some systems but not others. A factorial design with growth curves, followed by a resistance re-test after serial passage, addresses both effect and risk. Analyse the curves with replicate-level variance, for example area under the curve, rather than judging by eye. Related drug-action questions appear in pharmacology dissertation topics.

Microbiome and community profiling

35. How much does the choice of DNA extraction kit and 16S rRNA region change the composition reported for the same mock community?
A commercial mock community gives ground truth. Compare observed and expected abundances, report bias by taxon, since Gram-positive organisms are often under-recovered, and include a negative extraction control. Suits Master’s level in a laboratory with sequencing access.

36. Do differential abundance tools disagree on one published microbiome dataset, and which findings survive across methods?
Run ALDEx2, ANCOM-BC and a simple rarefaction-plus-test approach on one public case-control dataset, then report overlap in the taxa called significant. The central idea is compositionality: relative abundances cannot be compared as if they were counts. The statistical test selector helps with the final comparison tests.

37. How much of the bacterial signal reported for one low-biomass sample type is explained by reagent contamination?
Reanalysis with the R package decontam on public data that include negative controls shows how many reported taxa track DNA concentration or appear in blanks. The limit is that many older studies never sequenced controls, so choose those that did. Master’s or PhD level.

38. Does delivery mode still predict early-life Bifidobacterium abundance after accounting for antibiotic exposure and feeding method in published infant cohorts?
Public metagenomes allow a reanalysis, but metadata vary between studies. Build a variable dictionary before analysis, test study as a batch effect and avoid causal language. A clear option for undergraduates comfortable with R.

Genome-based projects with no wet laboratory

Public assemblies from NCBI and the European Nucleotide Archive support complete projects without a bench. Gene-level questions beyond microbial genomes belong with genetics dissertation topics. In every project below, describe sampling bias before comparing groups, because public genomes over-represent outbreaks and clinical isolates.

39. What plasmid and resistance gene content distinguishes Escherichia coli ST131 clade C genomes from other ST131 clades in public assemblies?
Screen assemblies with AMRFinderPlus or ResFinder and type plasmid replicons. Short-read assemblies fragment plasmids, so claims about plasmid carriage need caution. A Master’s-level project.

40. How often do hypervirulence markers and carbapenemase genes co-occur in public Klebsiella pneumoniae genomes, and has this changed by collection year?
Kleborate provides virulence and resistance scores that can be plotted by year and country. Collection year is often missing, so report how many genomes were excluded and why. Suits Master’s or PhD level.

41. How closely do CARD, ResFinder and AMRFinderPlus agree when applied to the same set of assemblies?
Disagreement between resistance databases changes what a study can claim about gene prevalence. Compare gene calls per assembly, check a sample of discrepancies by manual BLAST and state identity and coverage thresholds. The findings feed directly into any methods chapter that relies on one of these tools.

42. Where do mcr colistin resistance genes sit in public genomes: which plasmid types carry them, in which hosts, and how has that changed over time?
Plasmid replicon typing, for example IncX4 and IncI2, combined with host species and collection year, describes dissemination. Unless long-read data are used, conclusions about plasmid structure remain tentative.

43. Which mucA mutations recur in public Pseudomonas aeruginosa genomes from cystic fibrosis lung isolates, and do they co-occur with other alginate pathway changes?
mucA loss-of-function mutations are associated with the mucoid phenotype. Variant calling against the PAO1 reference needs careful handling of mixed populations and repeated isolates from one patient. Hypermutator status can also confound counts.

44. What does public genome data show about the M1UK lineage of Group A Streptococcus, and how does it align with UKHSA’s reports of seasonal activity?
The M1UK lineage was defined by lineage-specific SNPs and increased SpeA production, so public emm1 genomes can be classified with published SNP schemes. Match collection years to UKHSA’s updates on Group A streptococcal infections. Genome databases are not random samples, so avoid incidence claims. PhD chapter scale.

Critical appraisal and review projects

45. Do CRISPR-based diagnostic assays for respiratory pathogens keep their reported accuracy in studies run independently of the developers?
A diagnostic accuracy review using QUADAS-2 can compare developer-led and independent studies. Separate analytical sensitivity, meaning the limit of detection, from clinical sensitivity, because they answer different questions.

46. Does rapid pathogen identification from positive blood cultures shorten time to effective antibiotic therapy only when it is paired with stewardship review?
A systematic review that compares studies with and without a stewardship component tests this directly. Use consistent outcome definitions, such as time to effective therapy, mortality and length of stay, and check for publication bias.

Choosing between a laboratory project, a data project and a review

Each route carries a different risk. The table sets out what each needs and where it usually breaks.

Evidence route	What you need	Main risk
Own isolates or reference strains	Supervisor, containment, media, controls, bench time	Too few isolates, or contamination you cannot rule out
Residual diagnostic material	Laboratory agreement and the right approval route	Delay in approvals and uncertain access
Environmental sampling	Site permission, sampling plan, storage and transport	Pseudoreplication and weather-driven variation
Public surveillance tables	A defined question and a statistical method	Reading ascertainment changes as real change
Public sequences	Computing access, pipeline skills, careful metadata	Sampling bias and inconsistent metadata
Systematic or critical review	A protocol, search strategy and appraisal tool	A question too broad to screen in the time available

Whichever route you choose, the methods chapter has to defend it. Specialist dissertation methodology help can pressure-test a design before data collection starts. The sample size calculator is useful for any prevalence or comparison study.

Biosafety, ethics and containment: what decides feasibility

The HSE’s Approved List of biological agents sorts organisms into four hazard groups under the Control of Substances Hazardous to Health Regulations 2002. Some universities restrict student projects to hazard groups 1 and 2, and many treat containment level 2 work as supervised, trained-staff activity. If an organism is unlisted, your safety officer decides, and the higher group is the safe assumption when two seem possible.

Genetically modified microorganisms fall under separate contained-use rules, so a plasmid construct changes the paperwork. Human material adds another layer. If a project uses samples from NHS patients, expect either the NHS research ethics route or a service evaluation route that the laboratory confirms. Analysis of published aggregate tables is usually simpler, although your own committee makes that call. Ask in the first fortnight, because approvals commonly take longer than students expect.

Mistakes that cost marks in microbiology projects

• Counting specimens as isolates. If 30 per cent of specimens grow the target organism, 300 specimens give about 90 isolates, and fewer once repeat isolates from one patient are removed.
• Mixing breakpoint versions. Cite EUCAST or CLSI and the version year, and apply the same one to every isolate.
• Treating technical replicates as independent. Three plates from one culture are one biological replicate.
• Running a demonstration, not a study. Testing a plant extract against a standard strain and reporting a zone of inhibition shows something already expected. A comparison with a clear hypothesis, controls and a measured effect size is a study.
• Reading more reports as more disease. Testing, screening and notification rules change counts independently of incidence.
• Dropping controls. Negative extraction controls, sterility checks and quality control strains are part of the evidence, not extras.

Narrowing a broad microbiology dissertation idea

A broad idea such as “antibiotic resistance in hospitals” has no organism, no measurement and no method. Name five things instead: the organism or community, the setting, the measurement, the comparison and the time window. For example, that idea becomes topic 10: colistin disc diffusion error rates against broth microdilution in an archived Enterobacterales panel. The guide to choosing research topics covers the wider decision, and the dissertation topic finder can generate starting points in your own subfield.

Four topics turned into an aim and objectives

Topic 6, Candidozyma auris reporting.
Aim: to estimate how far the rise in reported C. auris cases in England between 2023 and 2025 reflects changes in screening and notification rather than transmission.

• To tabulate UKHSA case counts by region and specimen type, separating clinical infections from screening specimens.
• To compare the timing of increases in each region with the dates of guidance updates and of notification from 6 April 2025.
• To classify regions as consistent or inconsistent with a screening-only explanation, using stated ascertainment indicators.

Topic 10, colistin testing.
Aim: to measure the error rates of disc diffusion for colistin in Enterobacterales against broth microdilution.

• To test a panel of reference and archived isolates by both methods under one documented protocol.
• To calculate categorical agreement and very major, major and minor error rates using ISO 20776-2 definitions.
• To compare the results with EUCAST’s stated position on colistin testing.

Topic 18, bathing water.
Aim: to test whether storm overflow spill hours in the 72 hours before sampling explain E. coli exceedances at one named bathing water across four bathing seasons.

• To collect Environment Agency E. coli results and Event Duration Monitor data for overflows in the bathing water’s catchment.
• To build lagged spill variables at 24, 48 and 72 hours and add rainfall as a covariate.
• To estimate the effect of spill hours on exceedance probability, with confidence intervals.

Topic 40, Klebsiella pneumoniae genomes.
Aim: to describe the co-occurrence of hypervirulence markers and carbapenemase genes in public K. pneumoniae genomes by collection year.

• To assemble a dataset of public genomes using stated inclusion and exclusion criteria.
• To score each genome with Kleborate and quantify co-occurrence per year and region.
• To report excluded genomes and discuss sampling bias in the conclusions.

Matching microbiology topics to your level

Undergraduate projects suit one organism, one assay or one dataset, and a plain comparison. Topics 2, 5, 9, 17, 18, 26, 28, 38, 45 and 46 are written with that in mind. Master’s projects can handle method comparisons, environmental qPCR and genome analysis, for example topics 10, 13, 19, 36 and 39.

PhD work is different in kind. Topics 11, 22, 24, 30, 40 and 44 would usually form one chapter of a larger programme, and novelty has to be shown from the literature, not asserted. For any level, remember that a defensible small question beats an ambitious one that cannot be completed.

Practical checks before you commit

Can I complete a microbiology dissertation without laboratory access?

Yes. Topics 1 to 9, 17, 20, 23, 27, 32 and 36 to 46 use surveillance tables, published data or literature. Choose a method you can defend, such as time-series analysis, a systematic review protocol or a reproducible sequence pipeline, and state in the introduction why a desk-based design suits the question.

How many isolates or samples do I need?

That depends on the precision you want and the proportion you expect, so calculate it before collecting anything. Remember the denominator problem: you need the number of specimens required to yield enough isolates. The sample size calculator mentioned earlier handles the arithmetic, and a supervisor can confirm the assumptions.

Does a project using published surveillance tables need ethical approval?

It varies by institution. Many committees treat published aggregate tables differently from patient-level data, but the decision belongs to your university. Ask early, record the answer and mention it in the methods chapter.

Which susceptibility standard should a UK microbiology dissertation cite?

Most UK NHS laboratories report against EUCAST breakpoints, whereas CLSI is more common in North American and some other international data. Name the standard and the version year in your methods, because breakpoints change between versions and mixing standards across datasets distorts resistance percentages. If you run your own testing, state whether the method followed the ISO 20776 reference approach or a commercial system.

What containment level does a student microbiology project usually need?

Most taught dissertations use organisms that can be handled at containment level 1 or 2, such as non-pathogenic strains, environmental isolates or routine clinical-type bacteria under supervision. Hazard group 3 agents need containment level 3 facilities and are rarely suitable for a student project. The HSE Approved List classifies each agent, and your university biosafety officer decides what your department can host.

Which UK data sources can I use for an antimicrobial resistance dissertation?

Common starting points are the ESPAUR report tables, the UKHSA Fingertips antimicrobial resistance indicators and NHS prescribing data. Each uses different denominators and reporting years, so record the table, year and definition you used before you begin analysis. Furthermore, check whether a dataset reports isolates, patients or episodes, because the unit changes how you compare results.

What makes a microbiology dissertation topic narrow enough?

A narrow topic names one organism or community, one setting, one measurable outcome and one comparison. For example, resistance in Escherichia coli bloodstream isolates in one region across a defined period is workable, whereas antibiotic resistance in hospitals is not. If you cannot state the outcome as a number, a presence or absence result or a defined theme, the topic needs trimming.

From shortlist to research question

Pick two or three ideas, write the organism, setting, measurement and comparison for each, and bring them to your supervisor. Need help narrowing this down? The research question generator can turn a shortlisted idea into a testable question, and the research aims and objectives generator helps shape the next step. The guide on how to turn an idea into a research question explains the logic behind both.

If you want specialist feedback on a draft proposal before your supervisor meeting, research proposal support is available. Any reference material, from any source, is for learning and should be used in line with your institution’s policy.